Fundamentals of

Mobile
Computing
Operation Subsystem
(OSS)

Mobile Station Base Station Network and Switching

(MS) Subsystem (BSS) Subsystem (NSS)

Prasant Kumar Pattnaik

Rajib Mall
 FUNDAMENTALS OF
MOBILE COMPUTING

PRASANT KUMAR PATTNAIK

Associate Professor
School of Computer Engineering
KIIT University
Bhubaneswar

RAJIB MALL
Professor
Department of Computer Science and Engineering
Indian Institute of Technology Kharagpur

New Delhi-110001
2012
FUNDAMENTALS OF MOBILE COMPUTING
Prasant Kumar Pattnaik and Rajib Mall

© 2012 by PHI Learning Private Limited, New Delhi. All rights reserved. No part of this
book may be reproduced in any form, by mimeograph or any other means, without
permission in writing from the publisher.

ISBN-978-81-203-4632-1

The export rights of this book are vested solely with the publisher.

Published by Asoke K. Ghosh, PHI Learning Private Limited, M-97, Connaught Circus,
New Delhi-110001 and Printed by Raj Press, New Delhi-110012.
 Contents

Preface........................................................................................................................ix

Chapter 1:  Basics of Communication Technologies.............................. 1–23

1.1 Mobile Handsets, Wireless Communications, and
Server Applications...................................................................................... 1
1.2 Cell Phone System........................................................................................ 2
1.3 Types of Telecommunication Networks..................................................3
1.4 Computer Networks....................................................................................4
1.4.1 Controller Area Networks.............................................................4
1.4.2 Local Area Networks...................................................................... 5
1.4.3 Internetworks...................................................................................5
1.5 Traditional LAN............................................................................................ 6
1.6 LAN Architectures....................................................................................... 7
1.6.1 Bus Architecture.............................................................................. 7
1.6.2 Ring Architecture............................................................................8
1.7 Components of a Wireless Communication System............................. 9
1.8 Architecture of a Mobile Telecommunication System . ..................... 11
1.9 Wireless Networking Standards.............................................................. 12
1.10 Wireless Local Area Networks (WLANs).............................................. 15
1.10.1 Wireless LAN Architecture.........................................................16
1.10.2 Applications of Wireless LANs.................................................. 17
1.10.3 Advantages of Wireless LANs over Wired LANs.................18
1.11 Bluetooth Technology................................................................................ 19
1.11.1 Protocol Stack of Bluetooth......................................................... 20
Summary..................................................................................................................22
Further Readings...................................................................................................... 22
Exercises.................................................................................................................... 23

iii
iv  Contents

Chapter 2:  Introduction to Mobile Computing and Wireless

Networking...............................................................................24–46
2.1 What Is Mobile Computing?.................................................................... 24
2.2 Mobile Computing vs. Wireless Networking.......................................25
2.3 Mobile Computing Applications............................................................. 27
2.4 Characteristics of Mobile Computing ................................................... 27
2.5 Structure of Mobile Computing Application........................................28
2.6 Cellular Mobile Communication............................................................. 30
2.6.1 Generations of Cellular Communication Technologies.........31
2.7 Global System for Mobile Communications (GSM)............................35
2.7.1 GSM Services..................................................................................36
2.7.2 System Architecture of GSM....................................................... 37
2.7.3 GSM Security..................................................................................40
2.8 General Packet Radio Service (GPRS).................................................... 41
2.8.1 GPRS Services................................................................................ 41
2.8.2 GPRS Architecture.........................................................................41
2.9 Universal Mobile Telecommunications System (UMTS).................... 42
2.9.1 UMTS Network Architecture...................................................... 43
2.10 Mobile Phone and Human Body ...........................................................43
Summary..................................................................................................................44
Further Readings...................................................................................................... 45
Exercises.................................................................................................................... 45

Chapter 3:  MAC Protocols.........................................................................47–63

3.1 Properties Required of MAC Protocols.................................................47
3.2 Wireless MAC Protocols: Some Issues................................................... 48
3.2.1  The Hidden and Exposed Terminal Problems in
an Infrastructure-less Network....................................................48
3.3 A Taxonomy of MAC Protocols.............................................................. 50
3.4 Fixed Assignment Schemes......................................................................50
3.4.1 Frequency Division Multiple Access (FDMA).........................51
3.4.2 Time Division Multiple Access (TDMA).................................. 51
3.4.3 Code Division Multiple Access (CDMA).................................52
3.5 Random Assignment Schemes................................................................. 54
3.5.1 ALOHA Scheme............................................................................54
3.5.2 The CSMA Scheme....................................................................... 55
3.6 Reservation-based Schemes......................................................................55
3.6.1 MACA..............................................................................................56
3.7 The 802.11 MAC Standard........................................................................57
3.7.1 Infrastructure-based Mode........................................................... 57
3.8 MAC Protocols for Ad Hoc Networks..................................................58
Summary..................................................................................................................61
Further Readings...................................................................................................... 61
Exercises.................................................................................................................... 62
 Contents   v

Chapter 4:  Mobile Internet Protocol.......................................................64–77

4.1 Mobile IP...................................................................................................... 64
4.1.2 Terminologies—Mobile IP...........................................................66
4.2 Packet Delivery...........................................................................................68
4.3 Overview of Mobile IP.............................................................................. 68
4.4 Desirable Features of Mobile IP..............................................................70
4.5 Key Mechanism in Mobile IP.................................................................. 71
4.6 Route Optimization....................................................................................73
4.7 Dynamic Host Configuration Protocol (DHCP)................................... 74
4.7.1 Significance of Dynamic Host Configuration Protocol..........75
Summary..................................................................................................................75
Further Readings...................................................................................................... 76
Exercises.................................................................................................................... 76

Chapter 5:  Mobile Transport Layer.........................................................78–99

5.1 Overview of TCP/IP.................................................................................. 79
5.2 Terminologies of TCP/IP..........................................................................80
5.3 Architecture of TCP/IP............................................................................. 82
5.4 An Overview of the Operation of TCP.................................................84
5.5 Application Layer Protocols of TCP.......................................................86
5.6 TCP/IP versus ISO/OSI Model............................................................... 87
5.7 Adaptation of TCP Window....................................................................87
5.8 Improvement in TCP Performance.........................................................90
5.8.1 Traditional Networks.................................................................... 90
5.8.2 TCP in Mobile Networks............................................................. 92
5.8.3 TCP in Single-hop Wireless Networks.....................................92
5.8.4 TCP in Multi-hop Wireless Networks....................................... 96
Summary..................................................................................................................97
Further Readings...................................................................................................... 97
Exercises.................................................................................................................... 98

Chapter 6:  Mobile Databases................................................................ 100–115

6.1 Issues in Transaction Processing . ........................................................ 102
6.2 Transaction Processing Environment ..................................................103
6.2.1 Centralized Environment........................................................... 103
6.2.2 Client-server Environment......................................................... 104
6.2.3 Distributed Environment...........................................................104
6.2.4 Mobile Environment................................................................... 104
6.3 Data Dissemination.................................................................................. 107
6.4 Transaction Processing in Mobile Environment.................................108
6.4.1 Atomicity Relaxation..................................................................108
6.4.2 Consistency Relaxation...............................................................108
6.4.3 Isolation Relaxation..................................................................... 109
6.4.4 Durability Relaxation.................................................................. 109
vi  Contents

6.5 Data Replication........................................................................................109

6.6 Mobile Transaction Models.................................................................... 110
6.7 Rollback Process........................................................................................ 110
6.8 Two-phase Commit Protocol..................................................................110
6.9 Query Processing...................................................................................... 111
6.9.1 Location-dependent Querying..................................................111
6.9.2 Query Optimization.................................................................... 111
6.10 Recovery..................................................................................................... 113
Summary................................................................................................................113
Further Readings.................................................................................................... 114
Exercises.................................................................................................................. 114

Chapter 7:  Mobile Ad Hoc Networks................................................. 116–148

7.1 A Few Basics Concepts . ........................................................................117
7.1.1 How Is an Ad Hoc Network Set Up without
the Infrastructure Support?.......................................................117
7.1.2 Why Is Routing in a MANET a Complex Task?..................118
7.2 Characteristics of Mobile Ad Hoc Networks (MANETs).................118
7.2.1 MANET Operational Constraints ........................................... 120
7.3 Applications of MANETs........................................................................ 120
7.4 MANET Design Issues............................................................................ 122
7.5 Routing.......................................................................................................123
7.6 Essentials of Traditional Routing Protocols........................................124
7.6.1 Link State Protocols (LSP)......................................................... 125
7.6.2 Distance Vector (DV) Protocols .............................................. 127
7.7 Routing in MANETs: A Few Basic Concepts ................................... 128
7.7.1 Routing in MANETs vs. Routing in Traditional Networks...... 128
7.7.2 A Classification of Unicast MANET Routing Protocols......129
7.8 Popular MANET Routing Protocols.....................................................130
7.8.1 Destination-Sequenced Distance-Vector Routing Protocol.......131
7.8.2 Dynamic Source Routing (DSR) Protocol...............................132
7.8.3 Ad Hoc On-demand Distance Vector (AODV).....................135
7.8.4 Zone Routing Protocol............................................................... 135
7.8.5 Multicast Routing Protocols for MANET............................... 136
7.9 Vehicular Ad Hoc Networks (VANETs).............................................137
7.10 MANET vs. VANET................................................................................ 138
7.11  Security Issues in a MANET ................................................................ 138
7.12  Attacks on Ad Hoc Networks . ............................................................140
7.13  Security Attack Countermeasures......................................................... 143
Summary................................................................................................................144
Further Readings.................................................................................................... 144
Exercises.................................................................................................................. 146
 Contents   vii

Chapter 8:  Wireless Sensor Networks................................................ 149–165

8.1 WSN vs. MANET.....................................................................................150
8.2 Applications............................................................................................... 152
8.3 Architecture of the Sensor Node........................................................... 154
8.4 Challenges in the Design of an Effective WSN.................................155
8.5 Characteristics of Sensor Networks......................................................156
8.6 WSN Routing Protocols........................................................................... 157
8.6.1 Classification Based on Protocol Operation...........................157
8.6.2 Classification Based on Network Structure............................ 158
8.6.3 Directed Diffusion (DD).............................................................159
8.6.4 Rumor Routing............................................................................ 159
8.6.5 Sequential Assignment Routing (SAR)................................... 160
8.6.6 Low Energy Adaptive Clustering Hierarchy (LEACH) . ... 160
8.6.7 Power-Efficient Gathering in Sensor Information
Systems (PEGASIS).....................................................................160
8.6.8 Geographic and Energy Aware Routing (GEAR).................160
8.6.9 Geographic Adaptive Fidelity (GAF)...................................... 160
8.7 Target Coverage........................................................................................161
8.7.1 Some Strategies for Target Coverage...................................... 161
Summary................................................................................................................162
Further Readings.................................................................................................... 163
Exercises.................................................................................................................. 165

Chapter 9:  Operating Systems for Mobile Computing.................. 166–184

9.1 Operating System Responsibilites in Mobile Devices....................... 166
9.1.1 Managing Resources................................................................... 166
9.1.2 Providing Different Interfaces................................................... 167
9.2 Mobile O/S—A Few Basic Concepts...................................................167
9.3 Special Constraints and Requirements of Mobile O/S..................... 169
9.3.1 Special Constraints .................................................................... 169
9.3.2 Special Service Requirements .................................................. 171
9.4 A Survey of Commercial Mobile Operating Systems.......................172
9.4.1 Windows Mobile ........................................................................ 172
9.4.2 Palm OS......................................................................................... 174
9.4.3 Symbian OS..................................................................................175
9.4.4 iOS..................................................................................................177
9.4.5 Android . ......................................................................................177
9.4.6 Blackberry Operating System.................................................... 180
9.5 A Comparative Study of Mobile OSs................................................... 180
9.6 Operating Systems for Sensor Networks.............................................182
Summary................................................................................................................182
Further Readings.................................................................................................... 183
Exercises.................................................................................................................. 183
viii  Contents

Chapter 10:  Mobile Application Development and Protocols......185–198

10.1 Mobile Devices as Web Clients............................................................. 185
10.1.1 HDML (HanDheld Markup Language).................................. 187
10.2  WAP............................................................................................................ 188
10.3  J2ME............................................................................................................ 191
10.3.1 J2ME Configuration....................................................................192
10.4  Android Software Development Kit (SDK)........................................ 194
10.4.1 Android SDK Environment .....................................................194
10.4.2 Features of SDK...........................................................................195
10.4.3 Android Application Components...........................................195
10.4.4 Android Software Stack Structure........................................... 197
10.4.5 Advantages of Android.............................................................197
Summary................................................................................................................197
Exercises.................................................................................................................. 198

Chapter 11:  Mobile Commerce............................................................. 199–209

11.1 Applications of M-Commerce................................................................ 199
11.1.1 Business-to-Consumer (B2C) Applications............................. 200
11.2 Business-to-Business (B2B) Applications..............................................202
11.3 Structure of Mobile Commerce ............................................................203
11.4 Pros and Cons of M-Commerce............................................................205
11.5 Mobile Payment Systems........................................................................ 206
11.5.1 Mobile Payment Schemes.......................................................... 206
11.6 Security Issues........................................................................................... 208
Summary................................................................................................................208
Further Readings.................................................................................................... 209
Exercises.................................................................................................................. 209

Glossary........................................................................................................211–234

Index...............................................................................................................235–238
 Preface

This introductory text in the area of mobile computing is primarily based

on the classroom teachings of the first author. Mobile computing is
progressing at a rapid pace on account of the continual development of
new technologies and systems. Mobile computing is also included in the
curricula of several universities. The authors, therefore, decided to work
on an introductory textbook on this subject for the benefit of the students
and the teachers. The authors fervently hope that the readers will find
this book interesting as a whole and benefit from it by learning about the
fundamental range of topics covered. The authors in no way claim that the
material in the book is comprehensive.
The text is organized into eleven chapters and a glossary. Chapter 1
provides an overview of the basics of wireless technologies and computer
communications; it explains how a cellular communication system
functions—the technology that has revolutionized society.
Chapter 2 elucidates what mobile computing is and how it differs
from wireless networking. Several generations of cellular communication
technologies are discussed.
Chapter 3 describes how the medium access control (MAC) protocol
works in regulating the use of a shared physical channel among a set of
nodes. The chapter brings out, in essence, that protocol design for mobile
ad hoc networks is much more difficult and complex than that for wireless
LANs.
Chapter 4 provides an overview of the working of mobile internet
protocol (Mobile IP). Chapter 5 reviews some basic aspects of the TCP/IP
protocol suite. It brings out problems that might arise when TCP is used,
as it is, in mobile wireless networks. It then discusses the adaptations
that have been extended to make TCP work satisfactorily in the mobile
environment.
Chapter 6 on mobile databases explains how a mobile database is an
adaptation of a traditional database to accommodate many of the issues
ix
x  Preface

that a mobile environment imposes, such as disconnections, low battery

power, etc. A transaction model for a mobile computing application is also
discussed.
Chapter 7 is devoted to mobile ad hoc networks (MANETs). It discusses
routing protocols that have been proposed for use in MANETs, their
security vulnerabilities and the security measures incorporated at different
protocol layers. The routing protocols used in wireless sensor networks,
the challenges faced in the design and operation of wireless networks and
the various areas of their applications are all discussed in Chapter 8.
The features required of a mobile operating system, including
those available in popular commercial OSs, are discussed in Chapter 9.
Chapter 10 discusses J2ME and Android SDK that facilitate application
development for mobile phones.
Finally, in Chapter 11 an example of mobile computing infrastructure,
that is, mobile commerce (M-commerce) has been explained in simple
words.
The authors warmly welcome all readers to send their suggestions
for improvements to the book. All suggestions from readers towards
enhancements of the contents and the presentation style will be duly
acknowledged in the next edition of the book.
The first author is thankful to his wife Bismita and daughter
Prasannakshi, and the second author is thankful to his wife Prabina and
daughter Mithi for patiently and selflessly permitting them to devote time
towards the development of the book. Besides, both the authors profusely
thank their colleagues and friends for their cooperation and help during
the preparation of the manuscript.

Prasant Kumar Pattnaik

[email protected]

Rajib Mall
[email protected]
 Basics of Communication Technologies
1
I n this chapter, we review a few basic concepts related to telecommunica-
tions and computer communications that form the essential infrastructure
required for building knowledge in the area of mobile computing. In par-
ticular, we focus here on the basics of wireless technologies and computer
communications. These concepts form the prerequisites that we implicitly
assume in the treatment of this text.
Mobile computing refers to the computational tasks performed by
mobile users using their handsets. Since the handsets have very limited
processing power and memory, these devices by themselves do not have
the capability to carry out any significant and meaningful computations and
can only serve as the front-end for invoking remote applications. Mobile
computation, therefore, inevitably involves the invocation of applications
running on remote servers. In other words, mobile computation is
usually achieved by the interaction of a front-end application running
on the mobile handset with a server, seamlessly, through the medium of
wireless communication. Therefore, a study of mobile computation
involves the study of the invocation mechanisms at the client end, the
underlying wireless communication, and the corresponding server-side
technologies.

1.1  Mobile Handsets, Wireless Communications, and

    Server Applications

Currently the technologies forming the client-side (hand-helds) and the

server-side computations are in a state of flux and are rapidly evolving.
However, the wireless networking technologies have undergone significant
advancements over the last few decades and are nearly maturing.
Wireless networking has its roots in radio communications. In wireless
1
2  Fundamentals of Mobile Computing

communication, individual users with handsets may directly communicate

with each other over a radio link or via several radio and wired links
formed through some intermediaries such as base stations and fixed line
networks. When all the intermediaries are located on the ground, then
the communication system is referred to as a terrestrial radio system. If at
least one of the intermediaries is satellite borne, then it is referred to as a
satellite radio system.
Radio transmission was first discovered by Marconi in the year 1895,
that is, more than a century back. It would not be an overstatement to
say that since then the radio communication has revolutionized the entire
human civilization. Starting with the elementary, ground-breaking
point-to-point wireless communication achieved by Marconi, there has
been continually the induction of new and increasingly sophisticated
technologies. Sophisticated communication techniques such as conventional
radio, television, digital packet radio, cellular phone, wireless LAN, and
wireless-based Internet access have rapidly been developed over a relatively
short span of time. These have impacted our daily life, culture, and society
in no mean ways. More recently, a technology that has revolutionized the
human society and impacted almost everybody, is the cellular communication
technology.

1.2  Cell Phone System

Within a short span of less than two decades, more than half the world’s
population today owns cell phones. The cell phones have achieved
remarkable acceptance all over the world due to their major advantages over
the traditional phones. Besides providing the flexibility of communicating
while on the move, the cell phone system also provides data services such
as Short Message Service (SMS), Multimedia Messaging Service (MMS),
and even email and web browsing while on the move. Further, unlike a
traditional network, in a mobile telecommunication network, even when
the user moves, the current location of the mobile phone is maintained by
the network so that any voice call or SMS can be easily sent to the handset
wherever it may be. This has made it possible for people to communicate
and carry out important work anytime and from anywhere. For this reason,
mobile computing is often referred to as ubiquitous computing as well.
Today, we can say that cellular communication is influencing virtually
everyone’s life in some way or the other. Mobile computing is poised to
achieve higher levels of success in the current decade.
Before we proceed further, a brief narrative of how a cellular
communication system functions is presented in Box 1.1 to enable readers
to familiarize themselves with the architecture of the system.
 Basics of Communication Technologies   3

BOX 1.1  How a cellular communication system functions

Cellular telephony is based on radio frequency communications. A cellular
communication system divides the service areas into different geographic zones
called cells (see Fig. 1.4). Each cell can support a certain maximum number of
simultaneous user connections. Consequently, more cells are required per unit
area in cities, simply because of more people dwelling in cities and hence more
number of simultaneous calls occurring per unit area. Each user has a mobile
handset that has its own radio transmitter and receiver antenna to establish
connection to the local base station. The base station in turn connects to a
switching centre. The switching centre act as a middleman between two mobile
handsets and establishes connections between them when requested. As a user
moves from one place to another, his call is handled by the switching centre
of the base station located in another cell site, thus, providing a consistent,
high quality signal. When a user travels outside his home network, he can still
make calls. This is possible by a facility called roaming and it works because
the base station in the local cell is connected by wires with all other networks.
This makes it possible for a user to get his call almost anywhere. If someone
dials a user’s number, the network finds him and gives him the ring tone. We
will elaborate this narrative and give a more detailed overview of the cellular
communication architecture in Section 1.8.

1.3  Types of Telecommunication Networks

A popular way of classifying telecommunication networks is into voice

networks and data networks. The voice networks were the first ones to be
developed and formed the predominant telecommunication networks in
the nineteenth and twentieth centuries. In a voice network, analog traffic
is usually modulated on a carrier signal for spectral and transmission
efficiency. The traditional telephone networks that carry voice traffic in
analog form are an important example of a voice network. Data networks are
of much more recent origin. The data networks carry data in digital form.
The data signals are also modulated on an analog carrier signal for spectral
and transmission efficiency. The term data here refers to any information
such as text, documents, picture, movie, sound, etc., which needs to be
first coded into a bit stream. Of course, analog signals such as voice, video,
etc., have to be first quantized into digital data for coding into bit streams.
It is worth noting that the traditional voice networks such as PSTN
(public switched telephone networks) relied on circuit switching, whereas
data networks are more recent and are based on the store-and-forward
packet switching mechanism. A big advantage of the packet-switched
networks is that they make efficient use of the transmission medium
and therefore are much more cost effective compared to circuit-switched
networks. On the other hand, an inherent shortcoming of using a store-
and-forward network for digitized voice signal transmission is that in a
store-and-forward network, every switch inspects the packet to determine
4  Fundamentals of Mobile Computing

its destination and based on this information the network sends the packet
on its way forward.  Further, depending on the traffic conditions, packets
can get stored at a switch for a significant amount of time causing queues
to get built up at various switches. The cumulative delays that a packet
undergoes at various switches can be hundreds of times longer than the
propagation delay. So, the different packets of a voice call can undergo
different amounts of delay causing degraded sound quality. In a voice call,
even a moderate delay at a switch can cause crackling or popping sounds at
the receiver.   Unless the switches are built using very high-speed hardware,
the receiver of a call would hear such distracting noises. This was possibly a
reason why packet switching was not used for voice calls until recently. The
modern day networks deploy high-speed, powerful switches and routers.
This has made it possible for voice to be transmitted over packet-switched
networks and hence VoIP (Voice over Internet Protocol) has become a very
popular and cheap medium of communication. Computer networks are
predominantly based on the packet-switching technology.

1.4  Computer Networks

Several types of computer networks are in use today. Of these, we discuss

here Controller Area Networks (CANs), Local Area Networks (LANs), and
internetworks.

1.4.1  Controller Area Networks

A Controller Area Network (CAN) is essentially a very small network

that is typically used to connect the different components of an embedded
controller. The end-to-end length of a CAN is usually less than 50 metres.
Since the propagation time of a CAN is very small, it behaves more like
a local bus in a computer.
To understand the genesis of CAN and its operation, consider the
present-day automobiles. The area of automotive electronics has becomes
fairly sophisticated for providing support to several activities such as
engine management, fuel injection, active suspension, braking, lighting,
air conditioning, security and central locking. A considerable amount
of information exchange among the various automotive components
takes place when the engine is operational. The conventional method of
networking the components in older models of cars was the point-to-point
wiring of the different electrical components such as motor generator,
lamps, battery, ignition system, etc. As cars became more sophisticated,
the use of such a naive scheme would have required several kilometres of
wiring, adding not only to the cost of manufacturing but also contributing
to severely reduced reliability.
 Basics of Communication Technologies   5

The limitations of fixed point-to-point wiring techniques in handling

the demands of modern automated cars and other embedded applications,
gave rise to the development of CAN. A special requirement placed on
CAN is that it should be able to effectively handle noise. Automotive
components such as electric motors, ignition systems, as well as RF
transmissions, are heavy producers of noise. Another requirement imposed
on CAN is the use by it of the 12-volt power supply that was mandated
by the conventional 12-volt automotive power supply. CAN specifies only
the physical and data link layers of the ISO/OSI model while the higher
layers are left open for specific implementations.
Because of its robustness, the use of CAN has extended beyond its
automotive origins and can now be found in diverse application areas such
as industrial automation systems, trains, ships, agricultural machinery,
household appliances, office automation systems and elevators. Now CAN
is an international standard under ISO 11898 and ISO 11519–2.

1.4.2  Local Area Networks

A Local Area Network (LAN) is typically deployed in a building or a

campus and is usually privately owned. For example, a LAN can be used
to connect a number of computers within an organization to share data
and other resources such as files, printers, FAX services, etc. LANs typically
operate at data rates exceeding 10 Mbps and many present-day LANs
(gigabit Ethernets) operate at 1 Gbps.

1.4.3  Internetworks

Several LANs can be interconnected using switches to realize internetworks

or internet in short. Figure 1.1 shows four separate LANs interconnected
using switches to form an internet.

Figure 1.1  Structure of an internet.

6  Fundamentals of Mobile Computing

In an internet, a node in a LAN communicates with a node in another

LAN using packet switching. Packet switching implies that messages are
divided into relatively small units of data called packets at the sender’s
end before they are sent over the network. Each packet is transmitted
individually and traverses the network using the store-and-forward
principle. A packet can even follow a route different from that taken by
other packets of the same message to reach its destination. The individual
packets are routed through the network based on the destination address
contained within each packet. Once all the packets forming a message arrive
at the destination, they are ordered into the original message. The store-
and-forward principle followed in packet switching allows the same data
path to be shared among message transmissions from different sources.
The networks of this type are known as connectionless, in sharp contrast
to the dedicated connections realized in connection-oriented networks. The
two major packet-switching modes are, therefore, known as connectionless
(also known as datagram switching) and connection-oriented (also known
as virtual circuit switching). We do not discuss these here and any reader
unfamiliar with these can refer to a first-level computer networking book.
Wide area networks (WANs) are important examples of packet-switched
computer networks. The most prominent WAN is the Internet.
The Internet, sometimes referred to as “the net,” is a worldwide system
of computer networks—a network of networks in which users at any one
computer can, if they have permission, get information from any other
computer in the net using protocols such as ftp, http, etc. A user can even
talk directly to other users at other computers in the Internet using VoIP.
We must, however, be aware of the following two distinctions: the term
internet is used to denote a single logical network realized from a collection
of LANs by using an internetworking protocol, while the Internet refers to
the worldwide collection of interconnected networks which evolved from
the ARPANET project. It uses the Internet Protocol (IP) to link the various
physical networks spread all over the world into a single logical network.
The Internet began in 1962 as an attempt towards realizing a resilient
computer network for the US military and since then has grown into a
global network of more than a billion computers connected to several of
thousands of smaller computer networks, based on a common addressing
scheme called the IP address.

1.5  Traditional LAN

A traditional LAN architecture is schematically shown in Fig. 1.2. It is the

same as the bus architecture discussed in Section 1.6.1. There is a single
shared channel (bus) and only one node can transmit at any time. The
exact time when a node can transmit on the channel is determined by
the access arbitration policy of the network. The transmission control policy
 Basics of Communication Technologies   7

determines how long a node can transmit. These two policies together are
called the access control policy and form the Media Access Control (MAC)
layer protocol. In other words, the MAC protocol in a LAN consists of two
parts: an access arbitration part that determines when a node can use the
channel and the transmission control policy decides how long it can use
it, once it starts using it. Let us first review some basic aspects of LANs
that are crucial to understanding our subsequent discussions.

1.6  LAN Architectures

Two major LAN architectures are being used: the bus architecture and
the ring architecture. These two architectures use different access control
techniques. We first briefly discuss these two LAN architectures, and then
discuss the associated access control techniques.

1.6.1  Bus Architecture

The schematic representation of a bus-based network architecture is shown

in Fig. 1.2. In a bus-based architecture, nodes are connected to the network
cable using T-shaped network interface connectors. The terminating points
are placed at each end of the network cable. There is a single, shared
channel (bus) for which the transmitting nodes contend to gain access to
it. In a bus architecture, nodes communicate using broadcasting. The most
commonly used protocol for access control in traditional bus networks is the
Carrier Sense Multiple Access with Collision Detection (CSMA/CD). These
networks are also called multiple access networks. In these networks, when
two or more nodes transmit packets simultaneously, the transmissions
overlap in time and the resulting signal gets garbled. Such an event is
called a collision. A collision entails retransmission of the corrupted frame.

Figure 1.2  A traditional bus-based LAN.

In CSMA/CD networks, nodes continuously sense the channel to

determine whether the channel is idle. A node transmits a packet only
if it senses the channel to be idle. But, this does not guarantee that there
8  Fundamentals of Mobile Computing

will be no collisions. Several nodes might sense the channel to be idle at

the same time instant and start transmitting simultaneously, resulting in
a collision. Also, even after a node starts transmitting, another node may
not sense it till the signal has propagated to it. It should, therefore, be
clear that the larger the propagation delay of a network, the higher is the
probability of collisions of packets. Thus, a long LAN would suffer from
too many collisions and not be usable. This sets a limit for the maximum
LAN size.
Ethernet is a LAN standard based on the CSMA/CD media access
control protocol. CSMA/CD protocol does not define any specific
collision resolution mechanism. On the other hand, Ethernet uses the
Binary Exponential Back-off (BEB) algorithm for collision resolution. Due
to its ubiquity, high speed, simplicity and low cost, Ethernet has over
the years emerged as one of the most preferred LAN protocols. It is,
therefore, not surprising that many attempts have been made in the past to
develop protocols based on Ethernet to support real-time communication.
As far as the real-time communications are concerned, the logical ring
architecture possesses significant advantages over Ethernet due to its
inherent deterministic access arbitration mechanism in contrast to the
collision-based mechanism of Ethernet. In a collision-based network, under
high load situations the number of collisions per unit time would increase
very rapidly with load, leading to increased retransmissions, rapid drop in
throughput and rise in delay. As a result, in Ethernet the delay in message
transmission increases rapidly as the traffic increases.

1.6.2  Ring Architecture

In a ring network shown in Fig. 1.3, nodes are placed along the ring.
The nodes transmit in turn. Each node usually transmits for a certain
predetermined period of time. Therefore, packet transmission delays become
predictable and can be made sufficiently small as per requirement. As a
result, ring-based architectures are often preferred in real-time applications.

Figure 1.3  A ring network.

The ring architecture, however, suffers from a few important problems.

First, any break in the ring can bring the whole network down. This makes
 Basics of Communication Technologies   9

reliability of ring networks a major concern. Further, ring is a poor fit to the
linear topology normally found in many situations. This made researchers
to look for alternative technologies which can have the advantages of both
the bus and ring architectures. This led to the development of the token
bus architecture. A token bus is a bus-based architecture where the stations
on the bus are logically arranged in a ring with each station knowing the
address of the station to its “left” and “right”.
When the logical ring is initialized, the highest numbered station gets
a chance to start its transmission. After transmitting for a predetermined
duration, this station passes the transmission permission to its immediate
neighbour (left or right as per the convention adopted) by sending a special
control frame called a token. The token propagates around the logical ring.
At any time, only the token holder is permitted to transmit packets. Since
only one station at a time holds the token in a ring network, collisions
cannot occur. An important point that must be kept in mind is that the
physical order in which the stations are connected to the cable need not
be the same as the order of the stations in the logical ring. This is because
a cable is inherently a broadcast medium. Each station would receive
every frame transmitted, and discard those that are not addressed to it.
After a node exhausts it assigned time slot for transmission, it hands over
the token to its logical neighbour. For this, it transmits a special token,
specifically addressed to its logical neighbour in the ring, irrespective of
whether that station is physically adjacent to the concerned node or not.
The MAC protocol in a token ring network should support adding stations
to and removing stations from the logical ring. The protocols based on the
underlying ring architecture hold the advantage of being able to guarantee
the delivery time of a packet.

1.7  Components of a Wireless Communication System

A wireless communication system is built from various types of basic

components. The following are some of these basic types of components.

Transmitter:  The input to a wireless transmitter may be voice, video, data

or other types of signal meant to be transmitted to one or more distant
receivers. This signal is called the baseband signal. The basic function of
the transmitter is to modulate, or encode several baseband signals onto a
high frequency carrier signal. A modulated high frequency signal can be
radiated and propagated more effectively and helps make more efficient
use of the radio frequency (RF) spectrum than what the direct transmission
of the individual baseband signals can achieve.

Receiver:  The receiver receives the modulated signals and reverses the
functions of the transmitter component and thereby recovers the transmitted
10  Fundamentals of Mobile Computing

baseband signal. The antenna of the receiver is usually capable of receiving

the electromagnetic waves radiated from many sources over a relatively
broad frequency range.

Antenna:  The function of an antenna is to convert the electric signal from

a transmitter to a propagating electromagnetic RF wave; or conversely,
to convert a propagating RF wave to an electric signal in a receiver. In a
transceiver, a transmitter and a receiver are co-located for supporting full-
duplex communications. In this case, the same antenna is usually shared
by both the transmitter and the receiver. There are mainly two types
of antennas that are used on wireless networks: omni-directional and
directional. Omni-directional antennas can receive and transmit over 360
degrees. This can be compared to a light bulb that emits light all over. In
contrast, a directional antenna is similar to a flashlight, it focuses light in
some direction.

Filters:  Filters are a key component present in all wireless transmitters

and receivers. They are used to reject interfering signals lying outside the
operating band of receivers and transmitters. They also reject unwanted
noise signals generated by the amplifier circuitry.

Amplifiers:  An amplifier amplifies the strength (usually voltage) of a

signal. The important specifications of an amplifier include power gain
and the noise figure. The noise figure of an amplifier is a measure of how
much noise is added to the amplified signal by the amplifier circuitry. This
is most critical in the front-end of the receiver where the input signal is
very weak and it is desirable to minimize the noise added by the receiver
circuitry. Therefore, it is necessary that the first amplifier in the receiver
circuit has as low a noise figure as possible.

Mixers:  A mixer is typically used to achieve frequency conversion at the

transmitters and receivers. Frequency conversion is required because it is
advantageous to transmit signals at a higher frequency. This is achieved
by modulating a carrier waveform using the original baseband frequency.
When a baseband signal is mixed appropriately with a high frequency on a
carrier, it can be easily and efficiently radiated and becomes less susceptible
to noise and attenuation. Therefore, the transmission range increases and
the received signal quality improves. Further, multiple baseband signals
can be mixed with a carrier appropriately in order to efficiently utilize
the spectral bandwidth. This forms the essence of any signal modulation
technique. When multiple baseband signals modulate multiple carrier
frequencies and the different baseband signals are made to occupy non-
overlapping bandwidths over a frequency spectrum, a broadband signal is
obtained.
 Basics of Communication Technologies   11

1.8  Architecture of a Mobile Telecommunication System

A simplified architecture of a mobile telecommunication system has been

shown in Fig. 1.4. It has three main components: the core network, the
radio access network, and the mobile phones.

Figure 1.4  Architecture of a mobile telecommunication system.

Mobile handsets communicate over the radio access network. The

radio access network is primarily composed of the base stations which
communicate with the mobile phones using radio frequency electromagnetic
waves. As shown in Fig. 1.4, the coverage area is decomposed into
hexagonal cells. In each hexagonal cell, one base station is located. Two
types of radio channels are usually involved in the communication between
a base station and the cell phones: control channels and voice channels.
Control channels typically use frequency shift keying (FSK) and are used
for transferring control messages (data) between the mobile phone and
the base station. Voice channels typically use frequency modulation (FM).
A base station typically has two antennas of different characteristics. One
antenna is used for receiving and the other for transmitting. The use of
the two different types of antennas at the base station increases the ability
of the base station to receive the radio signal from mobiles that use very
low transmitter power levels. On the other hand, mobile handsets typically
use the same antenna for both receiving and transmitting.
The core network interconnects the base stations, switches the mobile
switching centre (MSC), and also provides an interface to other networks
such as the traditional telephone network (PSTN) and the Internet.
The interconnect used in the core network is required to provide high-
speed connectivity. Therefore, usually fibre optic cables are used as the
interconnect in the core network. But based on the terrain conditions,
microwave communication is also sometimes used. This interconnection
12  Fundamentals of Mobile Computing

in the core network must allow both voice and control information to be
exchanged between the switching system and the base station. The MSC
is connected to the landline telephone network to allow mobile telephones
to be connected to standard landline telephones. The core network is
responsible for transmitting voice calls, SMS (Short Message Service),
etc. from one phone to another through switches. The core network also
maintains a database that contains information about the subscribers and
the information about billing.

1.9  Wireless Networking Standards

Standardization is very important to the computer networking domain

since many protocols and devices need to interoperate in any practical
networking solution. Further, there can be various vendors manufacturing
the networking equipment. In the absence of appropriate standards, it
would become difficult to interoperate the products manufactured by
different vendors. Mainly, three international standardization bodies are
responsible for formulating the networking standards: ITU, IEEE and ISO.
The IEEE (Institute of Electrical and Electronics Engineers) is a non-
profit, technical professional association of members from over 150 countries.
It acts as a standards body. Standards are very important in networking
since multiple devices that are often heterogoneous and manufactured by
different vendors need to communicate. The IEEE proposes standards for
new technologies and maintains the old standards. The IEEE created the
802 group to help standardize the LAN technology. The 802.3 standard
from this group defines the requirements that a product must meet for it
to be considered “Ethernet”. Wireless Ethernet is defined by 802.11. The
802.11 standard is further broken down into more specific certifications,
such as 802.11a, 802.11b, and 802.11g. Each of these defines a different
method for providing wireless Ethernet. Each protocol specifies various
aspects of data transfer that distinguish them from the other protocols.
The 802.11 standards define rules for communication on wireless local
area networks (WLANs). The popular 802.11 standards include 802.11a,
802.11b and 802.11g. The 802.11 was the original standard in this family,
ratified in 1997. It defined WLANs that operated at 1–2 Mbps. This standard
is obsolete today, but its extensions are being used extensively.
Each extension to the original 802.11 appends a unique letter to its
name. For example, the standards 802.11a, 802.11b and 802.11g define
different types of signal modulation and frequencies of operation as shown
in Table 1.1.
The following IEEE 802.11 standards are being used for wireless local
area networking:
• 802.11a: 54 Mbps standard, 5 GHz signalling (ratified 1999)
• 802.11b: 11 Mbps standard, 2.4 GHz signalling (1999)
 Basics of Communication Technologies   13

TABLE 1.1  Wireless Networking Standards

Standard Data rate Information

IEEE 802.11 Up to 2 Mbps This specification has been extended into
in the 2.4 GHz 802.11b.
band
IEEE 802.11a Up to 54 Mbps Products that adhere to this standard are
(Wi-Fi) in the 5 GHz considered “Wi-Fi Certified.” Eight available
band channels. Less potential for RF interference
than 802.11b and 802.11g. Better than 802.11b
at supporting multimedia voice, video and
large-image applications in densely populated
user environments. Relatively shorter range
than 802.11b. Not interoperable with 802.11b.
IEEE 802.11b Up to 11 Mbps Products that adhere to this standard
(Wi-Fi) in the 2.4 GHz are considered “Wi-Fi Certified.” Not
band interoperable with 802.11a. Requires fewer
access points than 802.11a for coverage of
large areas. Offers high-speed access to data
at up to 300 feet from base station. 14 channels
available in the 2.4 GHz band (only 11 of
which can be used in the U.S. due to FCC
regulations) with only three non-overlapping
channels.
IEEE 802.11g Up to 54 Mbps Products that adhere to this standard are
(Wi-Fi) in the 2.4 GHz considered “Wi-Fi Certified.” May replace
band 802.11b. Improved security enhancements
over 802.11. Compatible with 802.11b.
14 channels available in the 2.4 GHz band
(only 11 of which can be used in the U.S.
due to FCC regulations) with only three non-
overlapping channels.
IEEE 802.16 Specifies Commonly referred to as WiMAX or less
(WiMAX) WiMAX in the commonly as WirelessMAN or the Air Interface
10 to 66 GHz Standard, IEEE 802.16 is a specification for
range fixed broadband wireless metropolitan access
networks (MANs)
IEEE 802.16a Added support Commonly referred to as WiMAX or less
(WiMAX) for the 2 to 11 commonly as WirelessMAN or the Air Interface
GHz range. Standard, IEEE 802.16 is a specification for
fixed broadband wireless metropolitan access
networks (MANs)
Bluetooth Up to 2 Mbps No native support for IP, so it does not support
in the 2.45 GHz TCP/IP and wireless LAN applications well.
band Not originally created to support wireless
LANs. Best suited for connecting PDAs, cell
phones and PCs in short intervals.
(Contd.)
14  Fundamentals of Mobile Computing

TABLE 1.1  Wireless Networking Standards (Contd.)

Standard Data rate Information
HiperLAN/1 Up to 20 Mbps Only in Europe. HiperLAN is totally ad-hoc,
(Europe) in the 5 GHz requiring no configuration and no central
band controller. Does not provide real isochronous
services. Relatively expensive to operate and
maintain. No guarantee of bandwidth.
HiperLAN/2 Up to 54 Mbps Only in Europe. Designed to carry ATM cells,
(Europe) in the 5 GHz IP packets, Firewire packets (IEEE 1394) and
band digital voice (from cellular phones). Better
quality of service than HiperLAN/1 and
guarantees bandwidth.
Open Air Pre-802.11 pro- OpenAir is the proprietary protocol from
tocol, using Proxim. All OpenAir products are based on
frequency Proxim’s module.
hopping and
0.8 and 1.6
Mbps bit rate

• 802.11c: Operation of bridge connections (moved to 802.1D)

• 802.11d: Worldwide compliance with regulations for use of wireless
signal spectrum (2001)
• 802.11e: Quality of Service (QoS) support (not yet ratified)
• 802.11F: Inter-Access Point Protocol recommendation for communi-
cation between access points to support roaming clients
(2003)
• 802.11g: 54 Mbps standard, 2.4 GHz signalling (2003)
• 802.11h: Enhanced version of 802.11a to support European
regulatory requirements (2003)
• 802.11i: Security improvements for the 802.11 family (2004)
• 802.11j: Enhancements to 5 GHz signalling to support the Japan
regulatory requirements (2004)
• 802.11k: WLAN system management (in progress)
• 802.11l: Skipped to avoid confusion with 802.11i
• 802.11m: Maintenance of 802.11 family documentation
• 802.11n: 100+ Mbps standard improvements over 802.11g (in progress)
• 802.11o: Skipped
• 802.11p: Wireless Access for the Vehicular Environment
• 802.11q: Skipped
• 802.11r: Fast roaming support via Basic Service Set transitions
• 802.11s: ESS mesh networking for access points
• 802.11T: Wireless Performance Prediction—recommendation for
testing standards and metrics
• 802.11u: Internetworking with 3G/cellular and other forms of
external networks
 Basics of Communication Technologies   15

• 802.11v: Wireless network management/device configuration

• 802.11w: Protected Management Frames security enhancement
• 802.11x: Skipped (generic name for the 802.11 family)
• 802.11y: Contention Based Protocol for interference avoidance.

1.10  Wireless Local Area Networks (WLANs)

A brief history of evolution of wireless networks is given in Box 1.2. Today,

Wireless Local Area Networks (WLANs) provide connectivity between
computers over short distances using the wireless medium. Typical indoor
applications of WLANs may be in educational institutes, office buildings
and factories where the required coverage distances are usually restricted
to less than a few hundred feet. In the absence of obstructions and with the
use of suitable antennas, ranges of up to a few kilometres can be obtained.
Wireless networks are especially useful when it is impossible or prohi-
bitively expensive to carry out wiring within or across buildings, or when
only temporary access is needed between computers. WLANs are useful to
provide connectivity among portable computers. As an example, consider
an educational institute where the students may carry their own laptop
and use those in classrooms, library or lounge as and when they require.

BOX 1.2  History of wireless networks

The concept of communication without wires is not new. Smoke signals and
tribal drums were used to communicate over short distances without cords or
wires. Eventually, communications over long distances became possible through
wires. Claude Chappe invented the telegraph in 1792 and Alexander Graham
Bell first sent voice transmissions over wire in 1876. However in 1894, near
Bologna, Italy, wireless communication was born. Gulielmo Marconi tapped
out a message, causing a bell to ring on the other end of the room without
using any wire. Scientists began searching for ways to broadcast speech using
Marconi’s wireless. In 1906, Reginald Fessenden did it by using amplitude
modulation. In 1935, the American engineer Edwin Armstrong introduced FM
(frequency modulation) radio waves, which used less power and achieved
reception of higher quality signals. On October 13, 1983, the first call on a
commercial cellular system was made in Chicago.

Most commercial WLAN products use a number of technologies in

accordance with the international standards specified by the Institute of
Electrical and Electronic Engineers (IEEE). These WLANs operate with
maximum data rates of up to 54 Mbps, which are much lower than the
data rates that can be achieved with the wired Ethernet connections.
WLANs almost universally use Internet Protocols (such as TCP/IP) for
communication between computers. These protocols are basically a set
of standard rules to facilitate communication between computers. The
16  Fundamentals of Mobile Computing

HIPERLAN standard provides for WLAN operation with bit rate up to

20 Mbps. This can be observed from Table 1.1. WLANs operate in specific
frequency ranges depending on the availability of spectrum in specific
countries.
A typical WLAN is comprised of a few important components. These
components are shown in Fig. 1.5 and are briefly discussed below.

Figure 1.5  Wireless LAN card, access point and bridge.

Access point:  It is a radio receiver/transmitter (also called transceiver) that

connects to the wired network. These are typically mounted on the roofs at
different locations of a building. You can spot them if you carefully observe
the roof of a building having wireless LAN. The transceiver exchanges
signals with the wireless LAN card in desktop or notebook PCs. A single
access point can support a small group of users. It is connected to a wired
network through cables and provides the connectivity between wireless
devices and the wired network.
Wireless LAN cards:  End-users access the WLAN through WLAN
adapters (wireless network interface cards) in their hand-helds. The LAN
card used to be mounted on the motherboard of a computer. Now, it is
inbuilt into the motherboards.
Bridge:  It is used for connecting two LANs that may be in two different
buildings or on two separate floors within the same building.

1.10.1  Wireless LAN Architecture

One access point can provide support to a small group of users and can
perform within a range up to a few hundreds of feet. The access point
(or the antenna attached to the access point) is usually mounted on roof
tops but may be mounted elsewhere, if it is practical to do so, as long as
the desired radio coverage is obtained. End users access the wireless LAN
through wireless-LAN adapters, which are integrated within the hand-
held computers. Wireless LAN adapters provide an interface between the
client’s network operating system and the air waves via an antenna.
 Basics of Communication Technologies   17

One of the main concerns of users of wireless LANs is the apparent

reduction in privacy and security. To address this issue, WLANs uses
multiple levels of security to prevent unauthorized access to network
resources. Figure 1.6 shows the architecture of an infrastructure-based IEEE
802.11 network.

Figure 1.6  Architecture of an infrastructure-based IEEE 802.11 network.

In Fig. 1.6, observe that the access points are connected through a
distribution system which usually uses a fibre optic cable. As shown,
mobile nodes are connected to access points. The 802.11 standard stipulates
that the distribution system may be of any technology, such as Ethernet,
token ring, or any other network type. The majority of actual installations,
however, utilize Ethernet (802.3). An 802.3-based distribution system (also
referred to as the “wired backbone”) consists of switches or hubs that
tie together users (PCs and access points) equipped with 802.3 network
interface cards (NICs). The switch or hub is somewhat analogous to an
802.11 access point. The main difference is that a hub or switch provides
connection over a physical medium and an access point provides wireless
connectivity. In the example of Fig. 1.6, two 802.11 LANs are connected via
a distribution system. A distribution system can also be used to increase
network coverage through roaming between cells. The distribution system
also provides connectivity to the Internet through a switch.

1.10.2  Applications of Wireless LANs

Applications of the wireless LAN are numerous. In the following, we

discuss only the four important application scenarios of WLANs.
18  Fundamentals of Mobile Computing

1. Campus wireless LANs:  Several organizations, hotels, retail outlets,

warehouses, factories, research centres, and educational institutions
are among many others who are recognizing the value of flexibility
and connectivity provided by wireless LANs. Users with laptops,
PDAs, palmtops can have access to the organization’s intranet and
to the Internet from any location in the campus. An effective way
to make the network-based services available to them is by creating
a wireless network.
2. Streamlining i nventory management:  Stock control and inventory
management are the important activities in almost every manufac-
turing organization and retail outlet. Traditionally, stockkeeping was
done manually, with the stockkeeper going around the godown and
noting down the stock levels, counting various item and tallying
them manually for inventory control. Human errors can lead to
stockkeeping mistakes resulting in various types of financial losses.
With wireless LAN, devices such as hand-held scanners, keypads
and bar code readers can be linked to database applications and
printers. Instead of being a laborious time consuming task, inventory
management becomes automatic, making stock and inventory control
more manageable.

3. Providing LAN facilities in difficult-to-wire areas:  Wireless LANs can

coexist with wired LANs and help in extending LAN connectivity
to locations where providing wired connections is difficult.
4. WLAN connectivity to geographically dispersed computers:  Networks
of wireless LANs are being used for providing high bandwidth
links to connect different branches of an organization spread over a
city. These links between WLANs are being set up as Metropolitan
Area Networks (MANs) for e-governance or for linking various
government organizations, information kiosks, public utility places,
and various branches of a private organization, etc.

1.10.3  Advantages of Wireless LANs over Wired LANs

With wireless LANs, users can access shared information without looking
for a place to plug in their network cable, and network managers can
set up networks. Wireless LANs offer the following advantages over the
traditional wired networks:
1. Mobility:  Wireless LAN systems can help users get information at
any place in their organization. Such types of support and service
are not possible with the wired networks.
2. Simplicity and speedy deployment:  The installation procedure of a
wireless LAN system is simple because it eliminates the need to
run wires through walls. A wireless LAN facility can be set up in
an area in a matter of few hours.
 Basics of Communication Technologies   19

3. Flexibility:  Wireless technology allows the network to be accessible

where wiring is difficult to lay. Consider an airport lounge, the
passengers can connect their computers to the network just sitting
at their seats.
4. Cost effectiveness:  While the initial investment required for wireless
LAN hardware can be higher than the cost of wired LAN hardware,
but in the long run the cost benefits of WLAN are reasonable because
of the environment which often requires frequent movements and
dynamic changes.

1.11  Bluetooth Technology

The Bluetooth technology has become a popular means to implement

a Personal Area Network (PAN). A PAN helps to interconnect a set of
computerized devices that an individual person might require. For example,
a PAN makes it possible to network various appliances used in daily life
such as a microwave oven, fridge, air conditioner, mobile phones, etc. This
connectivity makes some meaningful applications to become possible. For
example, the fridge might automatically send the list of items that are out
of stock to the mobile, which can remind a housewife when shopping in
the mall. In the present standard, using wireless PAN, a dynamic group
of less than 255 devices can be made to communicate within an area of
less than 10 metres in diameter.

BOX 1.3  How was bluetooth technology named!

The king of Denmark, Herald, was nicknamed Bluetooth since he was fond
of blueberries and as per legend, he ate so many of them that his teeth got
stained and turned blue. He was credited to have united Denmark and Norway
regions. The Bluetooth technology was named after the king, since it unifies
(that is, provides seamless connectivity) multiple devices.

A prominent advantage of Bluetooth is that it can be used to get rid of

the mesh of wires that are required for interconnecting various devices
that are positioned near each other. For example, the mouse, the camera,
the printer, etc. can all communicate with the computer using Bluetooth
connectivity, making it possible to get rid of the mesh of wires. Such a
Bluetooth connectivity among a set of devices is called a piconet. A piconet
is essentially a very small (pico means very small) network as illustrated
in Fig. 1.7.
A Bluetooth piconet is based on a master-slave communication
architecture. One of the computer-enabled devices is designated as the
master and the other devices become the slaves. In a piconet, one master
device can interconnect with up to seven active slave devices using
20  Fundamentals of Mobile Computing

Figure 1.7  A Bluetooth piconet.

Bluetooth to form an ad hoc network. In Bluetooth communication, the

slave units only respond to commands from the master. This allows the
Bluetooth MAC to be simple, efficient, and non-contention based. In a
piconet, data transfer rates of up to 2.1 Mbits are possible. A scatternet is
formed when an ad hoc network of more than one piconets is formed.

1.11.1  Protocol Stack of Bluetooth

Bluetooth protocol stack makes possible the communication of both

data and control among many devices in a PAN. The protocol stack is
schematically presented in Fig. 1.8. We first give a brief overview of the
protocol stack, and then give a more detailed description of a few specific
protocol layers.

Figure 1.8  Bluetooth protocol stack.

The lowermost layer is called the Radio Frequency (RF) layer. Radio
modem is specified in this layer. This link controls the packet but the bit
levels are specified by the baseband layer. The Link Management Protocol
(LMP) is capable of configuring links with other devices. It provides power
modes, traffic scheduling, packet format, authentication and encryption.
Logical Link Control and Adaptation Protocol (L2CAP) provide connection-
oriented and connectionless data services to the upper layers of the protocol
 Basics of Communication Technologies   21

stack. Control information may also be exchanged through it. L2CAP may
include services such as segmentation and re-assembly of data packets. The
Service Discovery Protocol (SDP) enables two or more Bluetooth devices
to support a particular service. In the following, we briefly describe the
functionalities of the different layers of the Bluetooth protocol stack.
Radio layer:  It defines the requirements on the Bluetooth transceiver
device that provides the required wireless support for master-slave
communications. It specifies the transceiver characteristics such as the
specific modulation technique used (GFSK, i.e. Gaussian Frequency Shift
Keying), interference performance, radio frequency tolerance, etc. The
Bluetooth radio is a low-power transceiver with a range of up to 10 metres
and uses spread spectrum transmission by frequency hopping.
Baseband:  The Baseband is the physical layer of the Bluetooth. As can be
seen in Fig. 1.8, this layer lies on top of the Bluetooth radio layer in the
Bluetooth stack. The functionalities of this layer include link establishment
and power control. It also defines the packet format and timing issues. The
baseband protocol implementation is called a Link Controller (LC).
Link Management Protocol (LMP):  LMP is a data link layer protocol
and is responsible for link setup between the Bluetooth devices. The
specific responsibilities that it carries out include link setup, authentication,
and link configuration. Two types of links are possible: Asynchronous
Connectionless Links (ACL) and Synchronous Connection-Oriented (SCO).
The ACL link is used for transmitting the packet-switched data received
from L2CAP. The SCO link is used for real-time data involved in voice
telephony applications.
Logical Link Control and Adaptation Protocol (L2CAP):  This layer
provides connection-oriented and connectionless data services to the upper
layers of the protocol stock. Using the services of this layer, the higher
level protocols can transmit and receive L2CAP messages that can be up
to 64 kilobytes in length. L2CAP forms the Bluetooth data packets through
message segmentation while transmitting and also forms messages from
the received Bluetooth data packets.

Services Discovery Protocol (SDP):  Each Bluetooth device maintains

an SDP server application that keeps track of the services available on
that device. By using this protocol, a mobile application can discover
which services are available on which device and can also determine the
characteristics of those available services. Thus, it becomes possible to
establish connection between two or more Bluetooth devices for specific
services. By using SDP, mobile applications can get information such as
(i) device information, (ii) services, and (iii) service characteristics.
Radio Frequency Communication (RFCOMM):  RFCOMM is a simple set
of transport protocols that are built on top of L2CAP, and helps realize
22  Fundamentals of Mobile Computing

TCP/IP connectivity. It also provides an emulated RS-232 serial port with

other Bluetooth devices, thus simplifying connectivity with many simple
and small devices without using wires.
Telephony Control Protocol Specification (TCS):  TCS defines the call
control signals for establishing a voice connection between Bluetooth
devices. This protocol also considers the mobility related issues of hand-
held devices. In order to avoid confusion with the popular transport layer
protocol, i.e., the Transmission Control Protocol (TCP), this telephone
control protocol specification is usually denoted by the abbreviation TCS.

SUMMARY

We first reviewed some of the basic concepts in computer networking

to provide the background necessary to understand the rest of this text.
Using traditional networking, people were unable to access information
on the networks or run remote application while they were moving. We
pointed out that wireless networking provides the flexibility of network
access to the mobile user, and over the last two decades or so it has
proved to be a powerful and indispensable technique for a general user.
However, several shortcomings such as the overall security of information,
disconnections, and degradation of signals as a user moves, need to be
adequately addressed. We briefly reviewed the important features of
WLANs, a popular means adopted today to provide wireless connectivity
to users.
While the advantages of wireless networking are numerous, possibly the
most severe handicap of wireless networking is the security of information.
The security issues arise from the basic broadcast nature of wireless
communication. Several techniques have been implemented to address the
security issues, which we will discuss in the subsequent chapters.

FURTHER READINGS

Cisco Systems, Inc., “Wireless Technologies.” Internetworking Technologies

Handbook (http://www.cisco.com/univercd/cc/td/doc/cisintwk/ito_
doc/wireless.pdf)
Joel Conover, “First Things First—Top 10 Things to Know About Wireless,”
Networking Computing, July 2000.
Rajib Mall, Real Time Systems: Theory and Practice, Pearson Education,
New Delhi, 2008.
Rhoton, John “The Wireless Internet Explained”, Digital Press, 2001.
 Basics of Communication Technologies   23

EXERCISES
1. What is the difference between a data network and a voice network?
What are their relative advantages?
2. What types of communication services are usually offered on data
and voice networks?
3. What are the main difficulties that would be experienced if digitized
voice signals are transmitted over a data network? How can these
difficulties be overcome?
4. What is Controller Area Network (CAN)? How is it different from
a LAN? What are the important applications of a CAN?
5. Explain why there is a bound on the length and the number of nodes
that can be connected to a LAN.
6. Determine the minimum packet size for a CSMA/CD network
operating at 1 Gbps and having a length of 100 metres. Clearly
show your assumptions and calculations.
7. What is a transceiver? What is the role of a transceiver in a wireless
communication network?
8. Explain the different hardware components that are used in a
wireless network.
9. What is wireless LAN (WLAN)? Explain the basic wireless LAN
architecture using a suitable schematic diagram.
10. What are the important advantages of a wireless LAN (WLAN) over
a traditional wired LAN?
11. Discuss the architecture of a mobile telecommunication network
using a suitable schematic diagram.
12. Briefly explain how a call can be set up between two mobile phones.
13. Why are standards necessary in networking? Briefly explain any
LAN standard.
14. What do you understand by signal modulation? Briefly explain how
it is achieved. Why is it necessary to modulate a baseband signal on
a carrier signal, before transmitting it?
15. What is the difference between a baseband signal and a broadband
signal? Using a suitable schematic diagram, explain how a broadband
signal is obtained.
16. What is Bluetooth? How is Bluetooth useful in mobile computing?
Describe the protocol stack of Bluetooth.
17. What is a personal area network? Explain the role of Bluetooth in
personal area networking using a suitable example.
18. What are piconets and scatternets? Mention some important
applications of these networks.
19. Discuss the architecture of a mobile telecommunication system using
a suitable schematic diagram.
 Introduction to Mobile Computing
2
and Wireless Networking

A few important breakthroughs achieved over the last few decades

have revolutionized the way people use computers. First, about two
decades ago the advancements made in the field of miniaturization of
electronic circuits made it possible to pack powerful processing units and
significant memory into portable laptops. More recently, the computing
elements have become more powerful and at the same time shrunk to fit
into palmtops that people can carry effortlessly. In another development,
high speed data communication facility has been made available to
portable computing platforms, largely through advancements in the areas
of computer communication and wireless networking technologies. These
two developments have contributed to the formation of the discipline of
mobile computing. This has made it possible for the users to take advantage
of a host of innovative services available in these platforms and use them
ubiquitously. As a result, people from all walks of life have started to carry
mobile handsets with them wherever they go and are able to perform
meaningful computations with them.

2.1  What Is Mobile Computing?

Mobile computing (sometimes called ubiquitous computing and also at times

called nomadic computing) is widely described as the ability to compute
remotely while on the move. This is a new and fast emerging discipline
that has made it possible for people to access information from anywhere
and at anytime. We can also view mobile computing as encompassing
two separate and distinct concepts: mobility and computing. Computing
denotes the capability to automatically carry out certain processing related
to service invocations on a remote computer. Mobility, on the other hand,
provides the capability to change location while communicating to invoke
24
 Introduction to Mobile Computing and Wireless Networking   25

computing services at some remote computers. The main advantage of this

type of mobile computing is the tremendous flexibility it provides to the
users. The user need not be tethered to the chair in front of his desktop, but
can move locally or even to far away places and at the same time achieve
what used to be performed while sitting in front of a desktop.

2.2  Mobile Computing vs. Wireless Networking

We must distinguish between mobile computing and wireless networking.

These two terms are not synonymous. While mobile computing essentially
denotes accessing information and remote computational services while
on the move, wireless networking provides the basic communication
infrastructure necessary to make this possible. Thus, we can say that
mobile computing is based on wireless networking and helps one to invoke
computing services on remote servers while on the move: be it be office,
home, conference, hotel, and so on.
It should be clear that wireless networking is an important ingredient
of mobile computing, but forms only one of the necessary ingredients
of mobile computing. Mobile computing also requires the applications
themselves—their design and development, and the hardware at the client
and server sides. In fact, we can say that mobile computing subsumes the
area of wireless networking. Consequently, to be able to understand the
subtle issues associated with mobile computing, in addition to studying
the different aspects of mobile computing applications, their design
and development, we need to have a good knowledge of the basics of
wireless communications technologies. Wireless networking is increasingly
replacing traditional networks because of the low setup time and low initial
investment required to set up the wireless network. As we discuss later in
this chapter, wireless networks appear in various forms such as WLANs
(Wireless LANs), mobile cellular networks, personal area networks (PANs),
and ad hoc networks, etc.
Wireless networks can be classified into two basic types. One is an
extension of wired networks. It uses fixed infrastructures such as base
stations to provide essentially single hop wireless communication with
a wired network as illustrated in Fig. 2.1 or a two-hop wireless cellular
communication with another mobile as explained earlier in Fig. 1.4.
The other type of wireless network is an ad hoc network. An ad hoc
network does not use any fixed infrastructure and is based on multi-hop
wireless communication as shown in Fig. 2.2.
One popular example of a fixed infrastructure wireless network is a
Wireless LAN (WLAN) that implements the IEEE 802.11 protocol. Observe
from Fig. 2.1 that only the last hop is through the wireless medium. An
access point (AP) provides the last hop connectivity of the mobile nodes
to a wired network. All communication goes through APs which perform
26  Fundamentals of Mobile Computing

Figure 2.1  Wireless network based on fixed infrastructures.

Figure 2.2  Wireless network having no fixed infrastructures.

bridging† between the wireless and the wired mediums. A station must be
recognized by an AP to be able to connect to the network. The AP may
require authentication and this in turn is used as the basic means to keep out
the unauthorized users. In an infrastructureless network, the communication
between hosts occurs directly or via a few intermediate nodes that form
the hops. For example, station A in Fig. 2.2 can communicate with station
C using either the hops A–B, B–C or A–D, D–C.
A recent development in this context, is wireless networking of various
types of devices using the Bluetooth technology. As already discussed in
Chapter 1, the Bluetooth technology can also be used to establish direct
wireless connection of cell phones with devices such as printers, cameras,
scanners, laptop and desk computers. Bluetooth is gradually replacing
cables and infrared as the dominant way of exchanging information
between devices. One of the objectives of the Bluetooth technology is to
enable users to easily connect to a wide range of personal computing and
telecommunication devices, without the need to buy, carry, or lay out
cables. In fact, the Bluetooth technology enables setting up of personal
area networks (PANs) known as piconets and ad hoc networks known
as scatternets. It provides opportunities for rapid deployment of ad hoc
connections, and the possibility of automatic, transparent connections
between devices. It promises to eliminate the need to purchase additional
or proprietary cabling and configuration exercises needed to connect the
individual devices.
†
A network bridge connects multiple network segments at the data link layer
(Layer 2) of the OSI reference model.
 Introduction to Mobile Computing and Wireless Networking   27

An ad hoc network is also known as a Mobile Ad hoc Network(MANET).

It is a collection of mobile nodes that form a network on the fly without
requiring the support of any fixed infrastructure. Wireless sensor networks
are a special type of wireless ad hoc networks. Mobile ad hoc networks
and sensor networks are discussed in more detail in Chapters 7 and 8
respectively.

2.3  Mobile Computing Applications

Mobile computing technology makes it possible for people to send or extract

information while on the move. For example, a stock broker travelling in a
car may wish to issue stock transaction orders from a mobile phone or to
receive share price quotations. As can be guessed, ease of deployment and
scalability are two important positive points in favour of data transmissions
over the wireless medium. But, it is not without some shortcomings. When
data is being transmitted on air, all the wireless devices present in the
transmission range can receive the data. This, therefore, opens up very
difficult security issues that must be overcome to ensure privacy of data.

2.4  Characteristics of Mobile Computing

A computing environment is said to be “mobile”, when either the sender

or the receiver of information can be on the move while transmitting
or receiving information. The following are some of the important
characteristics of a mobile computing environment.

Ubiquity:  The dictionary meaning of ubiquity is present everywhere. In

the context of mobile computing, ubiquity means the ability of a user to
perform computations from anywhere and at anytime. For example, a
business executive can receive business notifications and issue business
transactions as long he is in the wireless coverage area.

Location awareness:  A hand-held device equipped with global positioning

system (GPS) can transparently provide information about the current
location of a user to a tracking station. Many applications, ranging from
strategic to personalized services, require or get value additions by location-
based services For example, a person travelling by road in a car, may need
to find out a car maintenance service that may be available nearby. He can
easily locate such a service through mobile computing where an application
may show the nearby maintenance shop. A few other example applications
include traffic control, fleet management and emergency services. In a
traffic control application, the density of traffic along various roads can
be dynamically monitored, and traffic can be directed appropriately to
28  Fundamentals of Mobile Computing

reduce congestions. In a fleet management application, the manager of a

transport company can have up-to-date information regarding the position
of its fleet of vehicles, thus enabling him to plan accurately and provide
accurate information to customers regarding the state of their consignments.
Location awareness can also make emergency services more effective by
automatically directing the emergency service vehicles to the site of the call.

Adaptation:  Adaptation in the context of mobile computing implies

the ability of a system to adjust to bandwidth fluctuation without
inconveniencing the user. In a mobile computing environment, adaptation
is crucial because of intermittent disconnections and bandwidth fluctuations
that can arise due to a number of factors such as handoff, obstacles,
environmental noise, etc.

Broadcast:  Due to the broadcast nature of the underlying communication

network of a mobile computing environment, efficient delivery of data can
be made simultaneously to hundreds of mobile users. For example, all
users at a specific location, such as those near a railway station, may be
sent advertising information by a taxi service operator.

Personalization:  Services in a mobile environment can be easily

personalized according to a user’s profile. This is required to let the users
easily avail information with their hand-held devices. For example, a
mobile user may need only a certain type of information from specific
sources. This can be easily done through personalization.

2.5  Structure of Mobile Computing Application

A mobile computing application is usually structured in terms of the

functionalities implemented. The simple three-tier structure of a mobile
computing application is depicted in Fig. 2.3. Figure 2.4 shows a specific
scenario of the types of functionalities provided by each tier. As shown in
these figures, the three tiers are named presentation tier, application tier
and data tier.
We now briefly explain the roles of the three tiers of a mobile computing
application.

Presentation (Tier-1)

Application (Tier-2)

Data (Tier-3)

Figure 2.3  The three tier structure of a mobile computing application.

 Introduction to Mobile Computing and Wireless Networking   29

Figure 2.4  Functionalities provided by each tier structure of a mobile computing

application.

Presentation tier
The topmost level of a mobile computing application concerns the user
interface. A good user interface facilitates the users to issue requests and
to present the results to the them meaningfully. Obviously, the programs
at this layer run on the client’s computer. This layer usually includes web
browsers and customized client programs for dissemination of information
and for collection of data from the user.

Application tier
This layer has the vital responsibility of making logical decisions and
performing calculations. It also moves and processes data between the
presentation and data layers. We can consider the middle tier to be
like an “engine” of an automobile. It performs the processing of user
input, obtaining information and then making decisions. This layer is
implemented using technology like Java, .NET services, cold fusion, etc.
The implementation of this layer and the functionality provided by this
layer should be database independent. This layer of functionalities is
usually implemented on a fixed server.
30  Fundamentals of Mobile Computing

Data tier
The data tier is responsible for providing the basic facilities of data
storage, access, and manipulation. Often this layer contains a database.
The information is stored and retrieved from this database. But, when only
small amounts of data need to be stored, a file system can be used. This
layer is also implemented on a fixed server.

2.6  Cellular Mobile Communication

In a cellular mobile system, the area of coverage is split into cells as shown
in Fig. 2.5. Even though the cells have been shown to be non-overlapping
hexagons for simplicity, but in reality cell shapes are irregular and do
overlap to some extent. A base station (BS) is located at the centre of each
cell. The BS in a cell receives communications from all mobile handsets
in the cell and forwards the data to the appropriate handset. Thus, a
base station keeps track of the calls of all handsets in its cell. When a
mobile handset while still continuing a call, moves to another cell, the BS
“hands-off” the call to the BS in the new cell. When a cell covers a crowded
area having too many users, then the users can experience frequent call
drops. To overcome this problem, such cells are split into smaller cells.

Figure 2.5  Cell structure in a cellular mobile communication system.

Initially the focus of cellular mobile communication was voice

communication. But today cellular phones provide many services based on
data communication too. These include electronic mail, Internet access and
running a variety of mobile applications. The term mobile communication
has a much wider connotation than that of cellular communication, and
includes wireless LANs and ad hoc networks. However, due to the
overwhelming popularity of mobile phones, cellular communication and
mobile communication are at times used interchangeably.
 Introduction to Mobile Computing and Wireless Networking   31

2.6.1  Generations of Cellular Communication Technologies

Mobile communication technology has advanced at a very rapid pace

over the last five decades. The gradual technology improvements over the
last four decades can be roughly demarcated into four generations. Each
generation essentially provides higher data rate and additional capabilities,
as shown schematically in Fig. 2.6. This figure does not show the data rates
of technologies before GSM, since these were analog techniques that did
not support the data communications facility. The fourth generation (4G) of
technology provides a substantial order of magnitude improvements in data
speeds, but is not yet widely implemented. The important characteristics of
the various generations of cellular mobile systems have been summarized in
Table 2.1. As can be seen from the table, each passing generation of mobile
cellular system brought about significant advancements to the technology,
causing the quality of the services to improve and the number of service
offerings to increase, and at the same time the cost to the customer to drop
drastically. We briefly discuss these different generations of mobile cellular
communication systems in the following.

Figure 2.6  Summary of mobile technology advancements.

First generation
The first generation (1G) cellular system was designed in the late 1960s,
but was commercially deployed in the early 1980s. The first commercial 1G
system in the United States was known as Advanced Mobile Phone System
(AMPS). It became operational in 1982 and supported only voice calls.
This was a completely analog system. In an analog system, analog signals
are transmitted by modulating them on a higher frequency carrier signal,
without first converting the signal into digital form through quantization.
In a completely analog system, it is difficult to support SMS and other data
services. Also, the signals from different users cannot be intermixed on the
same channel, and have to be transmitted in clearly separated channels.
32  Fundamentals of Mobile Computing

TABLE 2.1  A Summary of the Important Characteristics of the Different

Generations of Cellular Wireless Communication Systems

Mobile
Features Standards Data speed
gen.
1G Analog transmissions, pri- NMT, AMPS, 600–1200 bps
marily supported voice TACS
communications.
2G Digital transmissions, impro- GSM 9.6 kbps
ved performance by letting
multiple users share a single
channel.
2.5G Enhanced multimedia and GPRS 28 kbps or
streaming video, web browsing. higher
3G Enhanced multimedia and UMTS, HSPDA, 384 kbps or
streaming video capabilities. EDGE, W-CDMA higher

In the 1G system, the available frequency spectrum was split into

a number of sub-bands (or channels), each of which was used by a
different caller. These systems typically allocated one 25 MHz frequency
band for the signals to be sent from the base station to the handset
(incoming signal), and a second different 25 MHz band for the signal
transmitted from the handset to the base station (outgoing signal).
Figure 2.7 shows a frequency band split into five sub-bands (channels).
Though for simplicity, we have shown the different channels to be adjacent
to each other, each channel was separated from the adjacent channels by a
spacing of about 30 kHz. This was called a guard band. The use of guard
bands was one of the causes of inefficient spectrum usage and resulted in
the reduced number of simultaneous calls that could be supported. This
problem was overcome in the subsequent generations of technologies.

Sub-band 1
Sub-band 2
Band Sub-band 3
Sub-band 4
Sub-band 5
Time

Figure 2.7  1G frequency band split into five sub-bands.

Different 1G standards were used in different countries:

• AMPS (Advanced Mobile Phone System) in the USA
• NMT 450 (Nordic Mobile Telephone) in various European countries
• TACS (Total Access Communications System) in the UK
 Introduction to Mobile Computing and Wireless Networking   33

The 1G systems were of multiple access type, since once a caller hanged
up, another caller could use the same frequency. For this reason, the 1G
technology was also called Frequency Division Multiple Access (FDMA).
It was possible to reuse the same frequencies in the non-adjacent cells,
because the transmitter power output was restricted. For example, the cells
shown shaded in Fig. 2.8 could use the same set of frequencies.

Figure 2.8  Cell structure and frequency allocation of 1G systems.

When a caller crossed a cell boundary, the channel being used might
not be made available in the new cell as it might already be in use in some
other cell. During handoff, a different channel was possible to be allocated
in this case, otherwise the call got dropped if none were available. Beside
the number of callers that could be accommodated being low, the voice
quality was poor due to analog transmission. Also, it provided no security
at all, since any one could hear a call by tuning into a channel.

Second generation
As already pointed out, the 1G technology had many disadvantages. The
major drawback was the small number of simultaneous calls that could
be made and the high risk of call drops during handoffs. Calls in 1G
were expensive because of the inherent inefficient usage of the bandwidth
spectrum and hence very few could afford to use a cell phone. Further, the
1G networks were not capable of providing several useful services such as
caller identity and SMS. The disadvantages of 1G systems were overcome
by the second generation (2G) cellular systems.
The 2G systems encoded voice and other information digitally before
transmitting them. Digital transmission has many advantages over analog
transmissions. These include noise immunity and better bandwidth utiliza-
tions. The 2G system offered significant advancements in the evolution of
the mobile cellular technologies. Hence the 2G technology rapidly replaced
the 1G technology because of the drastic reductions in the cost of phone
calls and availability of a wider range of services coupled with substantial
improvements in the quality of services. Also, SMSs became possible.
34  Fundamentals of Mobile Computing

However, we must remember that the 2G technology is in many

respects an extension of the 1G system and many of the principles involved
in a 1G system also apply to 2G. For example, they both use the same cell
structure. However, there are many differences. For example, they use
different signal modulation techniques. 2G uses CDMA (Code Division
Multiple Access) and TDMA (Time Division Multiple Access) as channel
access technology, while 1G used FDMA.
The 2G mobile system deployment started in the 1990s, and two
competing standards existed. In North America, the IS-95 standard was
adopted which used Code Division Multiple Access (CDMA) and could
multiplex up to 64 calls per channel in the 800 MHz band. In Europe and
elsewhere, operators adopted the Global System for Mobile Communication
(GSM) standard, which used Time Division Multiple Access (TDMA) to
multiplex up to 8 calls per channel in the 900 and 1800 MHz bands.
The first commercial deployment of Global System for Mobile
Communication (GSM) was done in 1992. It supported higher voice quality
and provided data services such as SMS and e-mail. We will discuss
the GSM system in more detail, later in this chapter. In 1993, another
2G system, known as CDMAone, was standardized and commercially
deployed in South Korea and Hong Kong in 1995, followed by the United
States of America in 1996.

2.5 Generation
General Packet Radio Service (GPRS) is an extension of GSM and is
considered to be the 2.5 generation technology. As indicated by the name,
it is based on packet switching compared to circuit switching used in 2G.
This was a significant improvement over 2G and helped to reduce call
costs dramatically. Another important advantage of GPRS is that it allows
users to remain connected to the Internet without incurring additional
charge and supports multimedia capabilities including graphics and video
communications. GPRS deployments began in 2000, followed by EDGE
in 2003. EDGE enhances the capabilities of GPRS, allows faster Internet
browsing, and makes it possible to use streaming applications. Though
this technology provided faster data rates over 2G systems, it is called
2.5G because it did not offer the multi-megabit data rates which are the
characteristics of the 3G system.

Third generation
The 3G systems are often referred to as IMT-2000 (International Mobile
Telecommunications-2000) systems since this was made a global standard
by ITU. The 3G systems support much higher data transmission rates and
offer increased bandwidth, which makes them suitable for high-speed
data applications as well as for high quality traditional voice calls. The
 Introduction to Mobile Computing and Wireless Networking   35

3G systems can be considered to be purely data networks, since voice

signals are converted to digital data, this results in speech being dealt
with in much the same way as any other form of data. The 3G systems
use packet-switching technology, and provide cheaper calls while giving
better average call quality than that of the 2G systems, but they do require
a somewhat different infrastructure compared to the 2G systems. The 3G
networks made it possible for service providers to offer many innovative
applications and services such as email, instant messaging and video
telephony, multimedia gaming, live-video buffering, and location-based
services among others. The first 3G network was deployed in Japan in
2001 by DoCoMo.
UMTS (Universal Mobile Telephone System) is one of the 3G mobile
systems that was developed within the ITU’s IMT-2000 framework. UMTS
was developed mainly for the GSM networks, so that these could be
easily upgraded to UMTS networks. In UMTS, coverage is provided by a
combination of a variety of cell sizes ranging from “in building” pico cells
to global cells provided by satellites.
Even though it was expected that the UMTS specification would become
a single global standard for 3G systems, it did not turn out that way. Now
many different versions of 3G systems have come into existence and each
one evolved from some existing 2G system. The main 3G technologies that
are prevalent include UMTS and CDMA2000. European countries have
adopted UMTS, while the USA uses CDMA2000.

Fourth generation
A 4G system provides a faster data rate than that of 3G (at least 10 times
faster) and makes mobile broadband Internet access possible. The 4G
system has made possible high speed Internet access from smartphones
and laptops with USB wireless modems. A few applications that could
not be supported in earlier generations of the cellular phone systems,
have now become possible in 4G including IP telephony, gaming services,
high-definition mobile TV, video conferencing and 3D television. The 4G
technology is expected to help solve the last mile problem that prevents
the mobile users from running applications that are available on wired
networks. There are at present two competing 4G standards: Mobile
WiMAX standard and the Long Term Evolution (LTE) standard.
In the following section, we provide a brief overview of the working of
a few cellular wireless technologies that are being popularly used at present.

2.7  Global System for Mobile Communications (GSM)

GSM (Global System for Mobile Communications) is at present being used
in India. It is possibly the most successful digital mobile system to have
ever been used till now. An important characteristic of the GSM system
36  Fundamentals of Mobile Computing

is that it provides data services in addition to voice services, and yet is

compatible to 1G systems.
GSM networks operate in four different radio frequencies. Most GSM
networks operate either in the 900 MHz or in the 1800 MHz frequency
bands. Some countries in the American continent (especially the USA and
Canada) use the 850 MHz and 1900 MHz bands because the 900 MHz and
1800 MHz frequency bands are already allocated for other purposes. The
relatively rarely used 400 MHz and 450 MHz frequency bands are assigned
in some countries, notably Scandinavia where these frequencies were
previously used for the first generation systems. In the 900 MHz band,
the uplink frequency band is 890–915 MHz, and the downlink frequency
band is 935–960 MHz.

2.7.1  GSM Services

GSM provides three main categories of services. These are:
(i) Bearer services
(ii) Teleservices
(iii) Supplementary services
In the following, we elaborate these different categories of services.

Bearer services
Bearer services give the subscribers the capability to send and receive data
to/from remote computers or mobile phones. For this reason, bearer services
are also known as data services (see Box 2.1). These services also enable the
transparent transmission of data between GSM and other networks like
PSTN, ISDN, etc. at rates from 300 bps to 9600 bps. These services are
implemented on the lower-three layers of the OSI reference model. Besides
supporting SMS, e-mail, voice mailbox, and Internet access, this service
provides the users with the capability to execute remote applications. GSM
supports data transfer rates of up to 9.6 kbps.

BOX 2.1  GSM bearer services

The GSM data services are named bearer services. Consider the following
example: Suppose a customer requires to send a data file such as a picture to
a computer at the office that is connected to a public telephone network. In
this example, the bearer service provides 9.6 kbps circuit-switched data transfer.
The handset dials the office computer telephone number and establishes a
connection with it via the modem. When the office computer modem accepts
the call, the customer’s handset begins to send data directly on the telephone
line channel at 9.6 kbps.

Bearer services permit either transparent or non-transparent, and either

synchronous or asynchronous modes of data transmission. We elaborate
these in the following.
 Introduction to Mobile Computing and Wireless Networking   37

• The transparent bearer services use the functions of the physical

layer of transmission of data leading to constant delay and
throughput if no transmission errors occur. There is a mechanism
called FEC (Forward Error Correction) to increase the quality of data
transmission.
• The non-transparent bearer services use protocols of the second and
third layers to implement error correction and flow control. They
use transparent bearer services in addition to a Radio Link Protocol
(RLP). This protocol comprises mechanisms of high level data link
control.

Teleservices
GSM provides both the voice-oriented teleservices and the non-voice
teleservices, as discussed below.

Telephony:  The main goal of GSM was to provide high quality digital
voice transmission, offering the bandwidth of 3.1 kHz of analog phone
systems. Special codecs are used for voice transmission, while other codecs
are used for the transmission of analog data for communication with
traditional computer modems used in fax machines.

Emergency number:  The same number is used throughout an area. This

service is free of cost and mandatorily provided by all service providers.
This connection will automatically be set up with the closest emergency
centre.

Short message services:  This service offers transmission of text messages

of sizes up to 160 characters. SMS services use the signalling channels,
making possible the duplex system of the sending and receiving the SMSs
messages.

Fax:  In this service, using modems fax data is transmitted as digital

data over the analog telephone network according to the ITU-T Standards
T.4 and T.30.

Supplementary services
GSM provides certain supplementary services such as user identification,
call redirection, and forwarding of ongoing calls. In addition, standard
ISDN features such as ‘close user groups’ and ‘multiparty’ communication
are available.

2.7.2  System Architecture of GSM

A GSM system consists of three main subsystems:
(i) Radio Subsystem (RSS)
38  Fundamentals of Mobile Computing

(ii) Networking and Switching Subsystem (NSS)

(iii) Operation Subsystem (OSS)
A schematic of the functional architecture of a GSM system is shown in
Fig. 2.9. The different components of this architecture are briefly explained
in the following.

Figure 2.9  Functional architecture of a GSM system.

Radio subsystem (RSS)

This subsystem comprises all the radio specific entities. That is, the mobile
stations, the base station subsystems, the base transceiver station and the
base station controller. We briefly explain the important components of
the radio subsystem in the following:

Mobile Station (MS):  A mobile station (MS) or cell phone contains two
major components: the subscriber identity module (SIM) and the mobile
device. The SIM is a removable smart card. Each mobile device has a
unique identifier that is known as its IMEI (International Mobile Equipment
Identity). Apart from the telephone interface, an MS also offers other types
of interfaces to the users such as USB, Bluetooth, etc. Despite its small size,
a SIM card is a very important component of a GSM network. It contains all
the subscription information of a subscriber and holds the key information
that activates the phone after it is powered on. It contains a microcontroller
to primarily store and retrieve data from the flash storage on the SIM.
Identification information is stored in the SIM card’s protected memory
(ROM) that is not accessible or modifiable by the customer. Additional
flash memory is included in the mobile device to allow storage of other
information such as addresses, pictures, audio and video clips, and short
 Introduction to Mobile Computing and Wireless Networking   39

messages. The SIM card contains many other identifiers and tables such
as card type, serial number, a list of subscribed services, and a Personal
Identity Number (PIN).

Base Station Subsystem (BSS):  A GSM network comprises many BSSs.

Each BSS consists of a Base Station Controller (BSC) and several Base
Transceiver Stations (BTSs). We will explain these components subsequently.
A BSS performs all functions necessary to maintain radio connections to
an MS, as well as does coding/decoding of voice.

Base Transceiver Station (BTS):  A BTS comprises all radio equipment

such as antenna, signal processors and amplifiers that are necessary for
radio transmission. It encodes the received signal, modulates it on a carrier
wave, and feeds the RF signals to the antenna. It communicates with both
the mobile station and the BSC.

Base Station Controller (BSC):  A BSC manages the radio resource of the
BTSs in the sense that it assigns frequency and time slots for all MSs in
the area. It also manages the handoff from one BTS to another within the
BSS. The BSC also multiplexes the radio channels onto the fixed network
connection to the Mobile Switching Centre (MSC).

Network and switching subsystem (NSS)

This subsystem forms the heart of the GSM system. It connects the wireless
networks to the standard public networks and carries out usage-based
charging, accounting, and also handles roaming. NSS consists of a switching
centre and several databases as described below.

Mobile Switching Center (MSC):  An MSC can be considered to form

the heart of a GSM network. An MSC sets up connections to other MSCs
and to other networks such as Public Data Network (PDN). An MSC is
responsible for the connection setup, connection release, and call handoff to
other MSCs. A Gateway MSC (GMSC) is responsible for gateway functions,
while a customer roams to other networks. It also performs certain other
supplementary services such as call forwarding, multiparty calls, etc.

Home Location Registers (HLRs):  A HLR stores in a database important

information that is specific to each subscriber. The information contains
subscriber’s IMSI, pre/post paid, user’s current location, etc.

Visitor Location Register (VLR):  It is essentially a temporary database that

is updated whenever a new MS enters its area by roaming. The information
is obtained from the corresponding HLR database. The function of the VLR
is to reduce the number of queries to the HLR and make the user feel as
if he were in his home network.
40  Fundamentals of Mobile Computing

Operation subsystem (OSS)

The operation subsystem contains all the functions necessary for network
operation and maintenance. It consists of the following:

Operation and Maintenance Centre (OMC):  It supervises all other

network entities. Its functions are traffic monitoring, subscribers, security
management and accounting billing.

Authentication Centre (AuC):  It protects against intruders targeting the

air interface. The AuC stores information concerned with security features
such as user authentication and encryption. The AuC is related to the HLR.

Equipment Identity Register (EIR):  It is essentially a database that is

used to track handsets using the IMEI. It helps to block calls from stolen,
unauthorized, or defective mobiles.

2.7.3  GSM Security

Security in GSM is broadly supported at three levels: operator’s level,

customer’s level and system level. These three levels help oversee aspects
such as correct billing to the customer, avoiding fraud, protecting services,
and ensuring anonymity. The following are a few important features
associated with providing security in GSM networks.

Authentication
The purpose of authentication is to protect the network against unauthorized
use. In the GSM context, it helps protect the GSM subscribers by denying
the possibility for intruders to impersonate authorized users. A GSM
network operator can verify the identity of the subscriber, making it highly
improbable to clone someone else’s mobile phone identity.
Authentication can be achieved in a simple way by using a password
such as Personal Identification Number (PIN). This method is not very
secure in GSM networks as an attacker can “listen” the PIN and easily
break the code.

Confidentiality
A GSM network protects voice, data and sensitive signalling information
(e.g. dialed digits) against eavesdropping on the radio path. Confidentiality
of subscriber-dialled information in the GSM network is achieved by
using encryption techniques prescribed by the GSM designers. Data on
the radio path is encrypted between the Mobile Equipment (ME) and
the BTS which protects user traffic and sensitive signalling data against
eavesdropping.
 Introduction to Mobile Computing and Wireless Networking   41

Anonymity
A GSM network protects against someone tracking the location of a user
or identifying calls made to (or from) the user by eavesdropping on the
radio path. The anonymity of the subscriber on the radio access link in
the GSM network is achieved by allocating Temporary Mobile Subscriber
Identity (TMSIs) instead of permanent identities. This helps to protect
against tracking a user’s location and obtaining information about a user’s
calling pattern.

2.8  General Packet Radio Service (GPRS)

GPRS when integrated with GSM, significantly improves and simplifies

Internet access. It transfers data packets from GSM mobile stations to
external packet data networks (PDNs). Packets can be directly routed from
the GPRS mobile stations to packet switched networks making it easy to
connect to the Internet.
GSM uses a billing system based on the time (duration) of connection,
whereas GPRS uses a billing system based on the amount of transmitted
data rather than the duration of the connection. So, users can remain
continuously connected to the system, and yet get charged only for the
amount of transmitted data.

2.8.1  GPRS Services

GPRS offers end-to-end packet-switched data transfer services which can

be categorized into the following two types:
(i) Point-to-Point (PTP) service
(ii) Point-to-Multipoint (PTM) service.
The PTP service is between two users and can either be connectionless or
connection-oriented. The PTM is a data transfer service from one user to
multiple users. Again, there are two types of PTM services. One is multicast
PTM where the data packets are broadcast in a certain area and the other is
group call PTM where the data packets are addressed to a group of users.

2.8.2  GPRS Architecture

GPRS architecture introduces two new network elements, called GPRS

Support Node (GSN) and the Gateway GPRS Support Node (GGSN).
A GSN is essentially a router. All GSNs are integrated into a standard
GSM architecture. The GGSN is the interworking unit between the GPRS
network and the external packet data network (PDN). The GGSN contains
42  Fundamentals of Mobile Computing

routing information for GPRS users, performs address connection and

tunnells data to a user through encapsulation. In Fig. 2.10, the GGSN is
connected to an external network and it transfers packets to the SGSN
through an IP-based GPRS backbone network.

Figure 2.10  GPRS architecture reference model.

As shown in Fig. 2.10, SGSN (Serving GPRS Support Node) helps

support MS. The SGSN is connected to BSC through frame relay and it is
at the same hierarchy level as the MSC. The GPRS Register (GR) is a part
of HLR which stores all the relevant GPRS data. In a part of HLR which
stores all the relevant data of GPRS in a mobile IP network, GGSN and
SGSNs can be compared with home agent and foreign agent respectively.
The data packets are transmitted to the BSS and finally to the MS through
the GGSN and SGSN. The MSC as we have already discussed is responsible
for data transport in the traditional circuit-switched GSM.

2.9  Universal Mobile Telecommunications System (UMTS)

CDMA2000 and UMTS were developed separately and are two separate
ITU approved 3G standards. In these networks, coverage is provided by
a combination of various cell sizes, ranging from “in building” pico cells
to global cells provided by satellites, giving service to the remote regions
of the world.
The UMTS was developed mainly for countries with GSM networks,
and it is expected that all GSM networks will be upgraded to UMTS
networks. Because it is a new technology, a whole new radio access
network has to be built. An important advantage of UMTS is that it gives
significantly enhanced capacities to operators.
 Introduction to Mobile Computing and Wireless Networking   43

The UMTS specification has been designed so that the UMTS systems
are compatible with GSM networks. Therefore, the UMTS networks can
easily work with any existing GSM/GPRS network. The UMTS systems use
different frequency bands, so the BTSs do not interfere with each other.
Let us now discuss the dissimilarities between these networks. The
UMTS networks are different from the 2G networks in the following
respects:

Higher speech quality:  In addition to speech traffic, the UMTS supports

the advanced data and information services and can be called a true
multimedia network.

Higher data rate:  The UMTS supports 2 Mbps data rate, which is much
higher than that supported by the 2G mobile systems.

Virtual home environment (VHE):  A user roaming from his network to

other UMTS networks will not feel any discontinuity or service difference,
thus giving a “feeling” of being in the home network. In contrast, in a 2G
network, a user is registered to a visitor location and is also charged a
roaming overhead.

2.9.1  UMTS Network Architecture

The UMTS network architecture can be divided into three main elements:

User Equipment (UE):  The User Equipment (UE) is the name by which
a cell phone is referred to. The new name was chosen because of the
considerably greater functionality that the UE incorporates compared to a
cell phone. It can be thought of as both a mobile phone used for talking
and a data terminal attached to a computer with no voice capability.

Radio Network Subsystem (RNS):  The RNS is the equivalent of the Base
Station Subsystem (BSS) in GSM. It provides and manages the wireless
interface for the overall network.

Core Network:  The core network is the equivalent of the GSM Network
Switching Subsystem (NSS).

2.10  Mobile Phone and Human Body

Extensive research has been conducted to study the possible health effects
of continuous exposure to low intensity RF fields produced by mobile
phones. The overall evidence suggests that mobile phone usage of less than
10 years does not pose any increased risk of brain tumour. The effect of
still longer use is unclear due to non-availability of data. Any conclusion
44  Fundamentals of Mobile Computing

therefore is uncertain and tentative. From the available data, however, it

does appear that there is no increased risk for brain tumours in long-term
users as well, with the exception of acoustic neuroma (see Box 2.2) for
which there is limited evidence of a weak association. Available studies
suggest that self-reported symptoms are not correlated to an acute exposure
to RF fields, but considering that the studies are carried out over a rather
limited duration, it is very difficult to draw any firm conclusions. Currently
available studies on neurological effects and reproductive effects have not
indicated any health risks at exposure levels below guidelines.

BOX 2.2  Acoustic neuroma

Acoustic neuroma is a non-cancerous tumour that develops on the nerve that
connects the ear to the brain. The tumour usually grows slowly. As it grows,
it presses against the hearing and balance nerves. At first, there may be no
symptoms or mild symptoms. They can include the following:
•  Loss of hearing on one side
•  Ringing in ears
•  Dizziness and balance problems

SUMMARY

Mobile computing is an umbrella term used to describe technologies

that enable people to access network services—anyplace, anytime, and
anywhere. The term ubiquitous computing and nomadic computing are
often used synonymously with mobile computing. Mobile computing
essentially involves accessing and manipulating information over a
wireless network using a hand-held device. The challenges in designing
effective mobile computing solutions include providing acceptable service
in the presence of low available bandwidth, intermittent disconnections,
poor security, and difficulty in assigning fixed addresses to hosts. Unlike
their wired counterparts, the design of software for mobile devices must
consider resource limitation, battery power and display size. Consequently,
new hardware and software techniques must be developed. For example,
applications need to be highly optimized for space, in order to fit in
the limited memory on the mobile devices. For Internet enabled devices,
the good old TCP/IP stack is not very suitable, it is computationally
expensive for hand-held devices, does not make efficient use of the available
bandwidth and is not optimized for low power consumption.
Digital cellular standards meet the current requirements of voice
communications and are being upgraded to meet the future demands in
mobile multimedia applications. The 3G mobile networks represent an
evolution in terms of capacity, data speeds and new service capabilities
from second generation mobile networks to provide an integrated solution
for mobile voice and data with wide area coverage.
 Introduction to Mobile Computing and Wireless Networking   45

FURTHER READINGS

Abdelsalam Helal, Bert Haskell, J. Carter, R. Brice, D. Woelk and

M. Rusinkiewicz, Anywhere Computing Concepts and Technology, Kluwer
Academic Publishers, September 1999.
Adelstein, Gupta, Rechard III, and Schwiebert, Fundamentals of Mobile and
Pervasive Computing, TMH.
Akyildiz, I.F. and J.S.M. Ho, “ Dynamic Mobile User Location Update for
Wireless PCS Networks”, ACM/Baltzer Wireless Networks Journal, 1995.
Arkady Zaslavsky and Zahir Tari, “Mobile Computing Overview and
Current Status”, Australian Computer Society, 1998.
Christian Bettstetter, Hans-Jorg Vogel, and Jorg Eberspacher, “GSM Phase 2
+ GPRS Architecture, Protocol and Air Interface”, IEEE Communication
Survey, 1999, Vol. 2.
Jan A. Audestad, “Network Aspect of GSM System”, EURCCON, June 1988.
Max Stepanov, “GSM Features and Security”, July 2004.
Rappaport, T., Wireless Communication and Principles, Pearson Education.
Tuan Huynh and Hoang Nguyen, “Overview of GSM and GSM Security”,
2003.

EXERCISES
1. State true or false against each of the following statements. Give
appropriate reasons for your answer.
(i) Mobile computing and wireless networking are synonymous
terms.
(ii) The Bluetooth technology is an example of an infrastructure-
less network.
(iii) WLAN is an example of an infrastructure-less network.
(iv) A sensor network is essentially based on a single hop wireless
communication.
(v) 2G cellular phones support electronic mail services.
(vi) In the Global Packet Radio Service (GPRS), the customer is
charged for Internet access based on connection time rather
than on total information download.
(vii) GSM stands for Group of Special Mobiles.
(viii) UMTS networks can easily work with the existing GSM/GPRS
networks.
2. What is mobile computing? Mention at least three applications of
mobile computing.
46  Fundamentals of Mobile Computing

3. Distinguish between mobile computing and wireless networking.

4. Give an overview of the working of current mobile cellular phones.
Briefly explain the distinguishing features of various generations of
wireless cellular networks.
5. Briefly explain how the mobile cellular communication has evolved
over different generations of technology.
6. What do you understand by 2.5G? Mention a few characteristic
features of this technology. How is it different from 2G and 3G
technologies?
7. Distinguish between infrastructure-based networks and infrastruc-
ture-less networks with the help of suitable schematic diagrams.
8. Briefly write about the Bluetooth technology. Give at least two
examples of its use.
9. Explain the architecture of a mobile computing environment. Define
the functions of the presentation tier, application tier and data tier
of mobile computing environment.
10. Briefly discuss the important functional differences between 1G, 2G,
and 3G cellular networks.
11. Compare 1G and 2G cellular wireless communication technologies.
12. Is 3G cellular wireless technology superior to 2G technology? Justify
your answer.
13. Explain how a GSM network provides security to the customers.
14. What are the advantages of GPRS over GSM?
15. What is UMTS? Describe the functions of HLR and VLR in call
routing and roaming?
16. Identify at least three similarities and three dissimilarities between
a GMS network and a UMTS network.
17. What do you mean by Virtual Home Environment (VHE)? Explain
how VHE is realized in 3G networks.
18. Do mobile phones affect the human body negatively? Explain your
answer.
19. Discuss the advantages and disadvantages of supporting TCP/IP in
a mobile computing network.
20. Identify the main reasons as to why a mobile handset is compact
and lightweight and yet provides a large number of features such
as roaming, camera, audio and video play and record, Internet
browsing, etc., while the traditional landline phone handsets are
bulky and provide only limited features.
21. What is the difference between analog and digital transmissions
in the context of mobile communications? What are their relative
advantages?
 MAC Protocols
3
I n a wireless network, multiple nodes may contend to transmit on the
same shared channel at the same time. In this situation, the transmitted
data would get garbled unless a suitable medium access arbitration scheme
is deployed. Usually, it is the responsibility of the medium access control
(MAC) protocol to perform this task. The MAC protocol is a sublayer of the
data link layer protocol and it directly invokes the physical layer protocol.
The primary responsibility of a MAC protocol is to enforce discipline
in the access of a shared channel when multiple nodes contend to access
that channel. At the same time, two other objectives of any MAC protocol
are maximization of the utilization of the channel and minimization of
average latency of transmission. However, a MAC protocol must be fair
and ensure that no node has to wait for an unduly long time, before it is
allowed to transmit. In the following, we identify the various characteristics
desirable of any MAC protocol.

3.1  Properties Required of MAC Protocols

The design of a MAC protocol depends upon the specific environment in

which it would be used and the specific design requirements to be met.
In spite of the wide variations in the characteristics of different protocols,
however, in a general sense a good MAC protocol needs to possess the
following features:
• It should implement some rules that help to enforce discipline when
multiple nodes contend for a shared channel.
• It should help maximize the utilization of the channel.
• Channel allocation needs to be fair. No node should be discriminated
against at any time and made to wait for an unduly long time for
transmission.
47
48  Fundamentals of Mobile Computing

• It should be capable of supporting several types of traffic having

different maximum and average bit rates.
• It should be robust in the face of equipment failures and changing
network conditions.
Many MAC layer protocols for wireless networks have already been
proposed, standardized, and are in use. Also, many other protocols that
work with improved efficiency or overcome some problems in specific
wireless environments, are being proposed by researchers and practitioners.
At present, IEEE 802.11 has emerged as a popular and standard MAC
protocol for wireless networks. IEEE 802.11-based network cards and
routers are available in the market that can be used to inexpensively and
easily set up wireless LANs (commonly referred to as Wi-fi hotspots). As
we discussed in Chapter 1, wireless networks can be divided mainly into
two categories: (a) infrastructure-based wireless networks that include
the WLANs, and (b) infrastructure-less wireless networks that include
the mobile ad hoc networks (MANETs). Though the MAC protocols for
these two environments have many things in common, MAC protocols for
Infrastructure-less networks are surprisingly much more complex as they
have to address certain additional problems that arise in the infrastructure-
less environments. We discuss MAC protocols for both these environments
in this chapter.
In Section 3.2, we first discuss a few basic issues concerning the MAC
protocols in a wireless network. In Section 3.3, we discuss a taxonomy of
MAC protocols for wireless networks. In the three subsequent sections,
we discuss the three important categories of MAC protocols: the fixed
assignment, the random access, and the reservation-based protocols.
Finally, we discuss the MAC protocols that have been specifically designed
for ad hoc networks.

3.2  Wireless MAC Protocols: Some Issues

A MAC protocol in a wireless medium is much more complex than its wired
counterpart. First, a collision detection scheme is difficult to implement
in a wireless environment, since collisions are hard to be detected by
the transmitting nodes. Also, in infrastructure-less networks, the issue of
hidden and exposed terminals make a MAC protocol extremely inefficient
unless special care is taken to overcome these problems. We elaborate the
hidden and exposed terminal problems in the following:

3.2.1  The Hidden and Exposed Terminal Problems in

an Infrastructure-less Network

The hidden terminal problem arises when at least three nodes (A, B, and C), as
shown in Fig. 3.1, communicate among each other. As shown in this figure,
 MAC Protocols   49

B is in the radio range of A, and B is also within the radio range of C.

However, the nodes A and C are not in the radio range of each other.
Note that if both A and C start to transmit to B at the same time, the data
received at node B would get garbled. Such a situation can arise because A
and C are “hidden” from each other, because they are outside each other’s
transmission range. In this situation, when one node starts to sense the
medium before transmission, it cannot sense that the other node is also
transmitting. This creates a very difficult and important arbitration problem
that a MAC protocol needs to resolve.

Figure 3.1  Hidden terminal problem.

A related problem called exposed terminal could arise in a scenario such

as that depicted in Fig. 3.2. MAC protocols usually inhibit transmission
when transmission from another terminal is detected. As a result, node
A will not be able to transmit to any node when B is transmitting to C.
On the other hand, had A transmitted to D, it would have been received
correctly by D and B’s transmission would have also been correctly received
at C. The problem arose only because A and B are within each other’s
transmission range, though the destination nodes are in the transmission
range of only one of the nodes. In other words, the problem occurs because
A is exposed to B’s transmission. The overall effect of this problem is that
it leads to inefficient spectrum usage as well as unnecessary transmission
delays unless these are carefully addressed by a wireless MAC protocol.

A B C

Figure 3.2  Exposed terminal problem.

50  Fundamentals of Mobile Computing

3.3  A Taxonomy of MAC Protocols

A large number of MAC protocols have been proposed. These MAC
protocols can be broadly divided into the following three categories:
(i) Fixed assignment schemes
(ii) Random assignment schemes
(iii) Reservation-based schemes
The fixed assignment schemes are usually called circuit-switched
schemes. In the fixed assignment schemes, the resources required for a
call are assigned for the entire duration of the call. On the other hand, the
random assignment schemes and the reservation schemes are called packet-
switched schemes. The random assignment schemes are comparable to the
connection-less packet-switching schemes. In this, no resource reservations
are made, the nodes simply start to transmit as soon as they have a packet
to send. In the reservation schemes, a node makes explicit reservation
of the channel for an entire call before transmitting. This is analogous
to a connection-based packet-switching scheme. The reservation-based
MAC schemes are suitable to handle calls with widely varying traffic
characteristics. In the following sections, we discuss these three categories
of MAC protocols in some more detail.

3.4  Fixed Assignment Schemes

A few important categories of fixed assignment MAC protocols are the
following:
• Frequency Division Multiple Access (FDMA)
• Time Division Multiple Access (TDMA)
• Code Division Multiple Access (CDMA)
We briefly discuss these techniques in the following subsections.

BOX 3.1  An analogy to the fixed assignment solution to the multiple

access issues of a shared medium
An analogy may be drawn to the fixed assignment solution to the multiple
access issues of a shared medium in the following way: Consider a students’
common room (channel) in which many students want to communicate with
each other. If the students want to avoid cross-talk in the ongoing process,
then either the students could take turns in speaking (i.e. time division), or
they could speak at different pitches (i.e. frequency division), or they could
speak in different languages (i.e. code division). The last analogy captures the
essence of CDMA, when the students who are speaking the same language
understand each other, but the rest of the students cannot. In CDMA, each
communicating pair shares a decryption code using which lets them understand
only the communication between them. In this case many codes occupy the
same channel, but only the users who share a specific code will be able to
understand each other.
 MAC Protocols   51

3.4.1  Frequency Division Multiple Access (FDMA)

In FDMA, the available bandwidth (frequency range) is divided into many
narrower frequency bands called channels. Figure 3.3 shows a division
of the existing bandwidth into many channels (showns as Ch 1, Ch 2,
etc.). For full duplex communication to take place, each user is allocated
a forward link (channel) for communicating from it (mobile handset) to
the base station (BS), and a reverse channel for communicating from the
BS to it. Thus, each user making a call is allocated two unique frequency
bands (channels), one for transmitting and the other for receiving signals
during the call. Obviously, when a call is underway, no other user would
be allocated the same frequency band to make a call. Unused transmission
time in a frequency band that occurs when the allocated caller pauses
between transmissions, or when no user is allocated a band, goes idle and
is wasted. FDMA, therefore, does not achieve a high channel utilization.

Figure 3.3  Channels in Frequency Division Multiple Access (FDMA) scheme.

3.4.2  Time Division Multiple Access (TDMA)

TDMA is an access method in which multiple nodes are allotted different
time slots to access the same physical channel. That is, the timeline is
divided into fixed-sized time slots and these are divided among multiple
nodes who can transmit. Note that in this case, all sources use the same
channel, but take turns in transmitting. Figure 3.4 shows the situation
where time slots are allocated to users in a round robin manner, with
each user being assigned one time slot per frame.  See Box 3.2. Obviously,
unused time slots go idle, leading to low channel utilization.

Figure 3.4  Channels in Time Division Multiple Access (TDMA) scheme.

BOX 3.2  TDMA scheme

In TDMA, each user of the channel owns the channel for exclusive use for one
time slot at a time in a round robin fashion.
52  Fundamentals of Mobile Computing

3.4.3  Code Division Multiple Access (CDMA)

In CDMA, multiple users are allotted different codes that consist of

sequences of 0 and 1 to access the same channel. As shown in Fig. 3.5, a
special coding scheme is used that allows signals from multiple users to be
multiplexed over the same physical channel. As shown in the figure, three
different users who have been assigned separate codes are multiplexed on
the same physical channel.

Figure 3.5  Schematic of operation of Code Division Multiple Access (CDMA).

In the following, we elaborate the CDMA technology. In CDMA,

multiple users use the same frequency at the same time and no time
scheduling is applied. All the senders send signals simultaneously through
a common medium. The bandwidth of this medium is much larger than
the space that would be allocated to each packet transmission during
FDMA and the signals can be distinguished from each other by means
of a special coding scheme that is used. This is done with the help of a
frequency spreading code known as the m-bit pseudo-noise (PN) code
sequence. Using m bits, 2m – 1 different codes can be obtained. From these
codes, each user will use only one code.

BOX 3.3  How to distinguish transmission from different nodes

Two vectors are said to be orthogonal if their inner product = 0. Let p and q
be two vectors and suppose p = (2, 5, 0) and q = (0, 0, 17), then the inner
product of pq = (20+50+017) = 0.
   For good autocorrelation, binary 0 is represented as –1 and binary 1 is
represented as +1. Let the binary sequence be 1001, then the representation
sequence is +1–1–1+1.
 MAC Protocols   53

It is possible to distinguish transmissions from different nodes by

ensuring some properties on the codes. A code for a user should be
orthogonal (that is, non-interfering) to the codes assigned to other nodes.
The term “orthogonal” means that the vector inner product is zero, and
good autocorrelation uses the bipolar notation where a code sequence of
binary 0 is represented as –1 and binary 1 is represented as +1. See Box 3.3.
On the receiving end, only the same PN sequence is able to demodulate
the signal to successfully convert the input data.

BOX 3.4  Pseudorandom sequence generator

To generate a series of pseudorandom numbers, a seed (or starting point) is
required. Based on the selected seed, the next number can be generated using a
deterministic mathematical transformation or can be generated probabilistically.
In CDMA, a code actually denotes a starting point (seed) for a pseudorandom
sequence generator (PRSG). PRSG generates a series of bits at a frequency
which is much higher than the actual user data (such as digitized voice). These
bits are XORd with the user data and subsequently the results are transmitted.
This occurs in the case of multiple transmitters.
   If someone listens to this signal with the help of a suitable wideband
receiver, the person will hear something similar to what is produced by random
noise. All the other users who are on the same frequency will send a similar
signal, but with a different PRSG seed. So these apparent random noises will
all coexist in the same band of frequencies, but would not interfere with each
other. This is due to the reason that the exact frequency of any transmitter at
any instant (which is in effect determined by the seed) is almost always unique.
Error correction takes care of occasional bit errors.
   The receiver is aware of the PRSG starting point for each transmitter. It
hears just one of the transmitters by correlating the noise it receives, against
its own PRSG, which is also running with the same seed. It is slightly similar
to FDMA in this sense, but the difference is that the transmitters do not
stay on one frequency. They hop around many times per bit of user data.
The pseudorandom sequence determines this hopping, rather than a fixed
assignment to each transmitter.

For simplicity, we assume that all nodes transmit on the same frequency
at the same time using the entire bandwidth of the transmission channel.
Each sender has a unique random number key, and the sender XORs the
signal with this random number key. The receiver can “tune” into this
signal if it knows the pseudorandom number. Consider an example, where
X, Y are the transmitters and Z is a receiver. Sender X_data = 1 and X_Key
= (010011). Its autocorrelation representation is (–1, +1, –1, –1, +1, +1). The
signal to be calculated at sender X is Xs = X_data  X_key = +1X_key =
(–1, +1, –1, –1, +1, +1). Similarly, sender Y_data = 0 and Y_key = (110101)
and the signal to be sent at Y is Ys = –1Y_key = –1(+1, +1, –1, +1, –1,
+1) = (–1, –1, +1, –1, +1, –1). The signal received by receiver Z is Xs + Ys
= (–1, +1, –1, –1, +1, +1) + (–1, –1, +1, –1, +1, –1) = (–2, 0, 0, –2, +2, 0).
At the receiver, in order to receive the data sent by sender X, the signal
54  Fundamentals of Mobile Computing

Z is dispread. So now if Z wants to get information of sender X data,

then ZX_key = (–2, 0, 0, –2, +2, 0)(–1, +1, –1, –1, +1, +1) = 2 + 0 + 0 +
2 + 2 + 0 = 6 > 0 (positive), that is the original bit was a 1. Similarly, the
information of sender Y data may be obtained as ZY_key = (–2, 0, 0, –2,
+2, 0)(+ 1, +1, –1, +1, –1, +1) = –2 + 0 + 0 – 2 – 2 + 0 = – 6 < 0 (negative).
So the Y data original bit was a 0.

3.5  Random Assignment Schemes

There are a number of random assignment schemes that are used in MAC
protocols. A few important ones are the following:
• ALOHA
• Slotted ALOHA
• CSMA
• CSMA/CD
• CSMA/CA

3.5.1  ALOHA Scheme

It is a simple communication scheme that was developed at the university

of Hawaii. The basic (also called pure) ALOHA scheme, is a simple
protocol. If a node has data to send, it begins to transmit. Note that the
first step implies that Pure ALOHA does not check whether the channel is
busy before transmitting. If the frame successfully reaches the destination
(receiver), the next frame is sent. If the frame fails to be received at the
destination, it is sent again. The simple ALOHA scheme works acceptably,
when the chances of contention are small (i.e., when a small number
of senders send data infrequently). However, the collisions can become
unacceptably high if the number of contenders for transmission is high.
An improvement over the pure ALOHA scheme is the slotted ALOHA.
In the slotted ALOHA scheme, the chances of collisions are attempted to
be reduced by enforcing the following restrictions. The time is divided
into equal-sized slots in which a packet can be sent. Thus, the size of the
packet is restricted. A node wanting to send a packet, can start to do so
only at the beginning of a slot. The slotted ALOHA system employs beacon
signals that are sent at precise intervals that mark the beginning of a slot,
at which point the nodes having data to send can start to transmit. Again,
this protocol does not work very well if the number of stations contending
to send data is high. In such cases, the CSMA scheme (described next)
works better.
 MAC Protocols   55

3.5.2  The CSMA Scheme

A popular MAC arbitration technique is the Carrier Sense Multiple Access

(CSMA). In this technique, a node senses the medium before starting to
transmit. If it senses that some transmission is already underway, it defers
its transmission. Two popular extensions of the basic CSMA technique are
the collision detection (CSMA/CD) and the collision avoidance (CSMA/
CA) techniques.
Unlike that in a wired network, in a wireless network the CSMA/CD
technique does not work very well. In the CSMA/CD technique, the sender
starts to transmit if it senses the channel to be free. But, even if it senses
the channel to be free, there can be a collision (why?) during transmission.
In a wired network, the implementation of a collision detection scheme is
simple. However, in a wireless network it is very difficult for a transmitting
node to detect a collision, since any received signal from other nodes
would be too feeble compared to its own signal and can easily be masked
by noise. As a result, a transmitting node would continue to transmit the
frame, and only the destination node would notice the corrupted frame
after it computes the checksum. This leads to retransmissions and severe
wastage of channel utilization. In contrast, in a wired network when a node
detects a collision, it immediately stops transmitting, thereby minimizing
channel wastage.
In a wireless network, a collision avoidance scheme works much better
compared to a collision detection-based scheme. A collision avoidance
scheme is based on the idea that it is necessary to prevent collisions at the
moment they are most likely to occur, that is, when the bus is released
after a packet transmission. We explain the reason for this in the following.
During the time a node is transmitting on the channel, several nodes might
be wanting to transmit. These nodes would be monitoring the channel
and waiting for it to become free. The moment the transmitting node
completes its transmission, these waiting nodes would sense the channel
to be free, and would all start transmitting at the same time. To overcome
such collisions, in the collision avoidance scheme, all nodes are forced to
wait for a random time and then sense the medium again, before starting
their transmission. If the medium is sensed to be busy, a node waiting to
transmit waits for a further random amount of time and so on. Thus, the
chance of two nodes starting to transmit at the same time would be greatly
reduced.

3.6  Reservation-based Schemes

A basic form of the reservation scheme is the RTS/CTS scheme. In an

RTS/CTS scheme, a sender transmits an RTS (Ready to Send) packet to
the receiver before the actual data transmission. On receiving this, the
56  Fundamentals of Mobile Computing

receiver sends a CTS (Clear to Send) packet, and the actual data transfer
commences only after that. When the other nodes sharing the medium
sense the CTS packet, they refrain from transmitting until the transmission
from the sending node is complete.
In a contention-based MAC protocol, a node wanting to send a message
first reserves the medium by using an appropriate control message. For
example, reservation of the medium can be achieved by transmitting a
“Ready To Send” (RTS) message and the corresponding destination node
accepting this request answers with a “Clear To Send” (CTS) message. Every
node that hears the RTS and CTS messages defers its transmission during
the specified time period in order to avoid a collision. A few examples of
RTS-CTS based MAC protocols are MACA, MACAW, MACA-BI, PAMAS,
DBTMA, MARCH, S-MAC protocols which have specifically been designed
for sensor networks. In the following, we discuss MACA as a representative
protocol belonging to this category of protocols.

3.6.1  MACA

MACA stands for Multiple Access Collision Avoidance. MACA solves the
hidden/exposed terminal problems by regulating the transmitter power.
A node running MACA requests to use the medium by sending an RTS
to the receiver. Since radio signals propagate omni-directionally, every
terminal within the sender’s radio range will hear this and then refrain
from transmitting. As soon as the receiver is ready to receive data, it
responds with a CTS.
Figure 3.6 schematically shows how MACA avoids the hidden terminal
problem. Before the start of its transmission, it sends a Request To Send
(RTS). B receives the RTS that contains the sender’s name and the receiver’s
name, as well as the length of the future transmission. In response to the
RTS, an acknowledgment from B is triggered indicating Clear To Send
(CTS). The CTS contains the names of the sender and receiver, and the
length of the planned transmission. This CTS is heard by C and the medium
is reserved for use by A for the duration of the transmission.

Figure 3.6  Hidden terminal solution in MACA.

 MAC Protocols   57

On receipt of a CTS from B, C refrains from transmitting anything for

the time indicated in the CTS. Thus a collision cannot occur at B during
data transmission, and the hidden terminal problem is solved.
Though this is a collision avoidance protocol, a collision can occur
during the sending of an RTS. Both A and C could send an RTS at same
time. But an RTS occurs over a very small duration compared to the
duration of data transmission. Thus the probability of collision remains
much less. B resolves this contention problem by acknowledging only one
station in the CTS. No transmission occurs without an appropriate CTS.
Figure 3.7 schematically shows how the exposed terminal problem is
solved in MACA. Assume that B needs to transmit to A. B has to transmit
an RTS first as shown in Fig. 3.7. The RTS would contain the names of the
receiver (A) and the sender (B). C does not act in response to this message
as it is not the receiver, but A responds with a CTS. C does not receive
this CTS and concludes that A is outside the detection range. Thus C can
start its transmission assuming that no collision would occur at A.

Figure 3.7  Exposed terminal solution in MACA.

3.7  The 802.11 MAC Standard

The IEEE 802.11 is the most widely used standard for WLANs today.
Low cost WLAN cards can be easily purchased from the market to set up
a WLAN. The IEEE 802.11 standard defines the functional aspects of the
medium access control (MAC) sublayer. The IEEE 802.11 defines separate
standards for infrastructure-based and ad hoc networks. The network
interface cards can be set to work in either of these two modes.

3.7.1  Infrastructure-based Mode

This mode of usage is widely used and helps provide Wi-Fi hot spots in a
campus to access the Internet. It is based on the CSMA/CA protocol due
to the obvious advantages of this protocol over the CSMA/CD schemes
58  Fundamentals of Mobile Computing

in a wireless environment, which we discussed in Section 3.5. In CSMA,

a station that wants to transmit will first have to listen to the channel for
a pre-determined period of time for checking if there is any transmission
activity occurring on the channel. The station would transmit if it senses the
channel to be idle. The transmission by the channel is deferred, if it senses
the channel to be busy. CSMA can be considered a greedy protocol since
a node trying to transmit, puts the packet out on the channel the moment
it finds that the channel is idle. CSMA uses collision avoidance to improve
the system’s performance by trying to be less greedy. In case, the channel
is sensed to be busy when it tries to transmit, then the transmission will
be deferred for a “random” interval, thereby reducing the probability of
further collisions on the channel. CSMA/CA is used in the 802.11-based
wireless LANs.
The RTS/CTS (Request to Send/Clear to Send) is an optional mechanism
used by the 802.11 to reduce frame collisions caused by the hidden node
problem. As we discussed, a Request to Send (RTS) packet is sent by a
sender and a Clear to Send (CTS) packet is sent by the intended receiver.
The CTS alerts all the nodes within the range of both the sender or the
receiver to keep quiet for the duration of the transmission.
The IEEE 802.11 has been extensively used because of its installation
and operational simplicity and low cost. However, it should be borne in
mind that it can be used only for a single-hop communication. That is, all
nodes can communicate as long as they are in the transmission range of a
single access point. This limitation can be overcome by multi-hop ad hoc
networking. The 802.11a operates at 54 Mbps. Note that in 802.11, since
the transmitter power is constant and limited, the data rate decreases when
the transmission range increases.

3.8  MAC Protocols for Ad Hoc Networks

Ad hoc networks are infrastructure-less, self-organizing networks of mobile

computers typically carried around by people. These computers are equipped
with radio-enabled network interface cards for communication purposes.
Every node can communicate directly with the other neighbouring nodes
who can “hear” its transmissions. The computers detect other computers
in their neighbourhood, dynamically forming links that set up the network.
A link is also called one “hop”. In addition to communicating with other
computers in the neighbourhood, a computer can communicate with remote
computers not in the vicinity by letting intermediate computers relay
information, thereby forming a multi-hop path to the required destination
computer. In forming a multi-hop path to a destination computer, a
complexity that arises is on account of the mobility of the computers.
Thus, a multi-hop path can become stale (that is, invalid) after some time,
or might even break while a communication is still going on. Thus, no fixed
infrastructure is required for setting up an ad hoc network.
 MAC Protocols   59

Wireless multi-hop ad hoc networks are comprised of lightweight

mobile radio frequency bands which are a scarce resource and therefore
these are shared among all hosts. Medium Access Control (MAC) protocols
for ad hoc networks can be either centralized or distributed. For instance,
Bluetooth uses a centralized scheme where the master of a network assigns
turns for transmission to slaves. The IEEE 802.11 is a MAC protocol that
can utilize both schemes, but in ad hoc networks the distributed scheme
is used. In this scheme, nodes contend for transmission turns, but if more
than one node wants to transmit at the same time, a random node is chosen
for transmission. The important categories of MAC protocols for MANET
are discussed below.
There are essentially two broad categories of MAC protocols for
ad hoc networks. The first category controls shared medium access by
letting terminals compete asynchronously. In this protocol, every node
is free to transmit any time. But upon detection of a collision, the access
arbitration policy is invoked to avoid the collision. The second type of
MAC protocol divides the medium into channels so that each competing
node uses a different channel, thereby avoiding collisions. One way to
achieve this is by dividing the transmission time (slots), and inserting a
frequency band between terminals and requiring them to synchronize in
order to ensure that they never use the same slot, frequency or code. In
ad hoc networks, only the terminals that are present in the neighbourhood
have to be synchronized. This is because the non-neighbour terminals
cannot interfere with each other. Garcia et. al have named the first type
“contention-based” protocols and the second type “contention-free”
protocols. The RTS-CTS based schemes are contention-based protocols and
the multiple access protocols (FDMA/TDMA/CDMA) are contention-free
protocols. In Fig. 3.8, we have shown this classification of the protocols.

Figure 3.8  A classification of MAC protocols for ad hoc networks.

In another category of protocols called fixed-assignment schemes, no

contention is involved. These protocols use techniques such as FDMA,TDMA
and CDMA. Protocols using this kind of medium access methods are
60  Fundamentals of Mobile Computing

contention-free because every node can transmit without worrying about

collisions.
As discussed, the RTS-CTS based schemes are contention-based.
Contention-based protocols incur additional overhead due to the
reservation messages. In a contention-based protocol, the delay that a
message undergoes is not deterministic, since it would depend on whether
the other nodes are trying to send messages at that time or not. Further, in
a contention-based protocol, as the network traffic increases, the number
of contentions increase. This in turn causes the throughput to degrade
dramatically. A positive aspect of the contention-based protocols is their
simplicity and relative easy of implementation.
Contention-free protocols overcome many of the shortcomings of the
contention-based protocols. Packets in a contention-free protocol are not
delayed on account of collisions. However, contention-free protocols are
difficult to implement in a multi-hop ad hoc network. We therefore do not
discuss contention-free ad hoc network protocols. In general, contention-free
protocols for wireless networks can be both centralized and distributed. For
instance, Bluetooth piconet uses a contention-free centralized MAC protocol
in which time slots are divided among terminals by the master. Also, cell-
based, single-hop, mobile phone networks use centralized contention-free
protocols. Interestingly, ad hoc networks that have controlling masters use
centralized contention-free protocols. Therefore, Bluetooth scatternet would
also qualify as a contention-free centralized MAC protocol.
Multi-access protocols do not need a centralized master but the
neighbouring nodes must agree on the specific resource (time, frequency, or
code) to use, so that there is no conflict with the other nodes. The problem
of assigning resource (time, frequency, code) to different pairs of nodes in
a multi-access protocol, is equivalent to the well-studied problem in graph
theory called the edge colouring problem in which no two edges (links) can
be assigned the same colour (out of a finite set of colours) if they share a
common vertex. We will not discuss the details of any MAC protocol and
restrict our discussions to the conceptual level only.
Much of the research done for MANET may also apply to sensor
networks since both operate as multi-hop wireless networks with power
constraints. MANET, however, focuses on device mobility, while sensor
networks normally have limited or no mobility. Long studied problems
in wireless networks, such as the hidden terminal and exposed node
problem, also exist in sensor networks, so protocol designers must handle
these issues in addition to the characteristics unique to sensor networks.
Researchers now have the challenge to solve existing problems from
traditional wireless networks under the constraints introduced by the
limited resources available in WSN. In this, nodes define common active/
sleep periods. The active periods are used for communication and the
sleep ones for saving energy. This approach requires that nodes maintain a
certain level of synchronization to keep the active/sleep periods common to
all nodes. During the active periods, nodes contend for the medium using
 MAC Protocols   61

CSMA, IEEE 802.11 DCF, etc. However, the use of common active/sleep
periods may not be suitable for applications with irregular traffic, because
nodes use contention inside active periods, which would be prohibitive
when nodes wake up without communicating, and may cause collisions
when there is high traffic that cannot be absorbed by the initially envisaged
size of the active periods.

SUMMARY

A MAC protocol regulates the use of a shared physical channel among a

set of nodes. It arbitrates among the contending nodes. We discussed the
three important categories of solutions: fixed-assignment, random access,
and reservation-based. Protocol design for mobile ad hoc networks is
much more complex than that for wireless LANs, since nodes in such
networks are constrained by factors such as low power, limited bandwidth,
intermittent link errors, and hidden and exposed terminal problems. The
IEEE 802.11 protocol has been implemented in low cost network cards and
access points, and has become widely accepted for the design of Wireless
LANs to provide Wi-Fi hot spots.

FURTHER READINGS

Deng, J. and Z. Haas, “Dual Busy Tone Multiple Access (DBTMA): A New
Medium Access Control for Packet Radio Networks”, Florence, Italy,
1998.
Dam, T.V. and K. Langendoen, “An Adaptive Energy-efficient MAC
Protocol for Wireless Sensor Networks”, in SenSys ’03, New York,
USA: ACM Press, pp. 171–180, 2003.
Heidemann, J., W. Ye, and D. Estrin, “Medium Access Control with
Co-ordinated Adaptive Sleeping for Wireless Sensor Networks”,
IEEE/ACM Transactions on Networking, 12, No. 3, pp. 493–506, 2004.
Karn, P., “A New Channel Access Method for Packet Radio”, ARRL/CRRL
Amateur Radio 9th Computer Networking Conference, Sept. 22, 1990.
Lichun Bao, J.J. Garcia-Luna-Aceves, “Hybrid Channel Access Scheduling in
Ad hoc Networks”, Proceedings of the 10th IEEE International Conference
on Network Protocols, 2002.
Rhee, I., A. Warrier, M. Aia, and J. Min, “ZMAC: a Hybrid MAC for
Wireless Sensor Networks”, New York, USA: ACM Press, 2005,
pp. 90–101.
62  Fundamentals of Mobile Computing

Rom, R. and M. Sidi, “Multiple Access Protocols—Performance and

Analysis”, Springer-Verlag, New York, 1990.
Singh, S. and C.S. Raghavendra, “PAMAS: Power-Aware Multi-Access
Protocol with Signalling for Ad hoc Networks”, ACM Computer
Communication Review, July 1998.
Sunil Kumar, Vineet S. Raghavan, and Jing Deng, “Medium Access Control
Protocols for Ad hoc Wireless Networks: A Survey”, Ad Hoc Networks,
Elsevier Science, Vol. 4, 2006, pp. 326–358.
Talucci, F., M. Gerla, and L. Fratta, “A Receiver Oriented Access Protocol
for Wireless Multihop Networks”, Proceedings of IEEE PIMRC, 1997.
Toh, C.K., Ad Hoc Wireless Mobile Wireless Networks, Prentice Hall PTR, 2002.
Vaduvur Bharghavan, Alan Demers, Scott Shenker, and Lixia Zhang,
“MACAW: A Media Access Protocol for Wireless LANs”, ACM
SIGCOMM Computer Communication, Review, Vol. 24, Issue 4, October
1994.

EXERCISES
1. What is the role of a MAC protocol? At which ISO/OSI layer does
it operate?
2. What is a hidden terminal? What problem does it create during
wireless communications? Explain your answer using a suitable
schematic diagram.
3. When does the exposed terminal problem arise? Explain your answer
using a suitable example.
4. What are the principal responsibilities of the MAC protocols? How
do MAC protocols for wireless networks differ from those in wired
networks?
5. What are the broad categories of MAC protocols? Name one popular
protocol from each of these categories.
6. Explain the working of a contention-based MAC protocol. Give two
examples of contention-based MAC protocols.
7. What are the different categories of MAC protocols. Identify the
situations under which protocols from one category would be
preferable over the other categories. Explain the working of a
reservation-based MAC protocol.
8. Why are collision-detection based protocols not suitable for wireless
networks?
9. Name one MAC protocol that is used in mobile ad hoc networks.
Briefly explain its working.
 MAC Protocols   63

10. Name one MAC protocol that is used in sensor networks. Briefly
explain its working.
11. What is MACA protocol? In which environment is it suitable? Briefly
explain its working. How does MACA protocol solve the hidden/
exposed terminal problems?
12. What do you mean by a schedule-based MAC protocol? Name a
schedule-based MAC protocol. Briefly explain its working.
13. What is FDMA? Briefly explain its working and at least one of its
important applications.
14. What is TDMA? Briefly explain its working and at least one of its
important applications.
15. Explain the basic scheme of the CDMA protocol. What is the role of
a pseudorandom sequence generator in the working of the CDMA
protocol?
16. Briefly explain the IEEE 802.11 standard and discuss its application.
17. Do you agree with the following statement: “In CSMA/CD protocol,
when two nodes transmit on a shared medium, a collision can occur
only when two nodes start transmitting exactly at the same time
instant.” Explain your answer.
18. Identify the specific reasons as to why the MAC protocols designed
for infrastructure-based wireless networks may not work satisfactorily
in infrastructure-less environments.
 Mobile Internet Protocol
4
T he Internet is built on top of a collection of protocols, called the
TCP/IP protocol suite. Transmission Control Protocol (TCP) and
Internet Protocol (IP) are the core protocols in this suite. IP is responsible for
routing a packet to any host, connected to the Internet, uniquely identified
by an assigned IP address. This raises one of the most vexing issues caused
by host mobility. In the traditional IP addressing scheme, each LAN is
assigned an address. The nodes in the LAN are assigned an address based
on the LAN address. In the traditional IP addressing scheme, when a host
moves to a different location, it may move to another network. As a result,
it needs to change its IP address. This is an unworkable proposition for
routing messages to a host, as it would keep changing its address as it
moves from one network to another.
In this context, a Mobile Internet Protocol (Mobile IP) was proposed by
the Internet Engineering Task Force (IETF). The mobile IP allows mobile
computers to stay connected to the Internet regardless of their location and
without changing their IP address. In other words, Mobile IP is a standard
protocol that extends the Internet Protocol by making mobility transparent
to applications and to higher level protocols like TCP.
Every mobile user likes to have a continuous network connectivity
irrespective of its physical location. The traditional IP does not support
user mobility. Mobile IP was created by extending IP to enable users to
keep the same IP address while travelling to a different network. Box 4.1
explains how the necessity of mobile IP arose with the introduction of
mobile computers to the Internet.

4.1  Mobile IP

The packet delivery to and from a mobile node has been schematically
shown in Fig. 4.1. In this figure, a correspondent node (CN) is connected
64
 Mobile Internet Protocol   65

BOX 4.1  Evolution of mobile IP

The IP defined in RFC 791 is the widely-used version of the Internet Protocol.
Interestingly, however, the present form of IP is not version 1 of IP but version
4. The TCP/IP suite evolved through three earlier versions, and was split into
TCP and IP layers in version 4. So, when you use IP today, you are referring
to IP version 4 (IPv4). It is therefore usual that “IP” means “IP version 4.
Despite the careful design of IPv4, it was later recognized that IP suffers
from several shortcomings. For example, it would not be able to support the
enormous number of users that are expected to use Internet in a couple of
years. Also IP does not distinguish among the different applications, and treats
all applications equally. However, the quality of service (QoS) requirement of
different applications may be different. For example, a streaming video requires
that video frames be transmitted without delay jitters, whereas applications
such as e-mail can tolerate considerable delay. This needed the development of
a new version of IP. In fact, a new version of IP has already been developed. It
is formally called Internet Protocol version 6 (IPv6) and also sometimes referred
to as IP Next Generation or IPng.
   The Internet was started at a time when mobile computers did not exist. As
a result, the present Internet lacks support for mobile users travelling across
the world. To address this issues, mobile IP was proposed. Mobile IP (MIP)
supports mobility at Internet Layer (Network Layer).

via a router to the Internet, and the home network and the foreign network
are also connected via a router, i.e. the home agent (HA) and foreign
agent (FA), respectively, to the Internet. Therefore, home agent (HA) is
implemented on the router connecting the home network with the Internet,
a foreign agent (FA) is also implemented on the router connecting the
foreign network with the Internet. The tunnel for the packets towards the
mobile node starts at the home agent and ends at the foreign agent, again
here the foreign agent has the care-of-address (COA).

Figure 4.1  Packet delivery to and from a mobile node.

It can be observed in Fig. 4.1 that several terminologies are associated

with mobile IP. We explain these terminologies in the following
subsection.
66  Fundamentals of Mobile Computing

4.1.2  Terminologies—Mobile IP

Various terminologies associated with mobile IP and used in Fig. 4.1 are
explained below:
Mobile Node (MN):  A mobile node is a hand-held equipment with
roaming capabilities. It can be a cell phone, personal digital assistant,
laptop, etc.
Home Network:  The home network of a mobile device is the network
within which the device receives its identifying IP address (home address).
In other words, a home network is a subnet to which a mobile node belongs
to as per its assigned IP address. Within the home network, there is no
need of mobile IP.
Home Address (HA):  The home address of a mobile device is the IP
address assigned to the device within its home network.The IP address on
the current network is known as home address.
Foreign Agent (FA):  The foreign agent is a router in a foreign network
that functions as the point of attachment for a mobile node when it roams
to the foreign network. The packets from the home agent are sent to the
foreign node which delivers it to the mobile node.
Foreign Network:  The foreign network is the current subnet to which the
mobile node is visiting. It is different from home network. In other words,
a foreign network is the network in which a mobile node is operating when
away from its home network.
Correspondent Node (CN):  The home agent is a router on the home network
serving as the anchor point for communication with the mobile node. It
tunnells packets from a device on the Internet, called a correspondent node
(CN), to the roaming mobile node.

BOX 4.2  Tunnelling process

The packet is forwarded by the home agent to the foreign agent. When the
packet comes to the foreign agent (care-of-address), it delivers the packet to
the mobile node. This process is called tunnelling. Tunnelling has two primary
functions: encapsulation of the data packet to reach the tunnel endpoint, and
decapsulation when the packet is delivered at that endpoint.

Care-of-Address (COA):  It is the address that is used to identify the

present location of a foreign agent. The packets sent to the MN are delivered
to COA.
The COA can be any of the following two types:
(a) Foreign agent COA:  The COA is an IP address of foreign agent (FA).
(b) Co-located COA:  When the mobile node (MN) acquires a temporary
IP address, that address acts as the COA.
 Mobile Internet Protocol   67

BOX 4.3  Care-of-Address (COA)

In real life, if a person is not in his own house and living in a temporary location,
that location is called his care-of-address (C/O) but, here, the care- of-address
defines the current location of the mobile node from an IP point of view. All
IP packets sent to mobile nodes are delivered to the care-of-address (COA),
i.e. not directly to the IP address of the mobile node.

Note:  The co-located address (temporary IP address) can be acquired

using services like dynamic host configuration protocol (DHCP).

Home Agent (HA):  It is located in home network and it provides several

services for the MN. HA maintains a location registry. The location registry
keeps track of the node locations using the current care-of-address of the
MN.

Agent Discovery:  During call establishment it is necessary for a mobile

node to determine its foreign agent. This task is referred to as agent
discovery. The following two discovery methods are popularly used:
(1) Agent advertisement, and (2) Agent solicitation.
In the following, we briefly explain these two agent discovery methods.
1. Agent advertisement:  Generally the foreign and the home agents
advertise their presence through periodic agent advertisement
messages. An agent advertisement message, lists one or more care-
of-addresses and a flag indicating whether it is a home agent or a
foreign agent. Agent advertisement is a popularly used method in
agent discovery.
2. Agent solicitation:  In case a mobile node (MN) does not receive any
COA, then the MN should send an agent solicitation message. But
it is important to monitor that these agent solicitation messages do
not flood the network. A mobile node can usually send up to three
solicitation messages (one per second) as soon as it enters a new
network. The basic purpose of the solicitation messages sent by a
mobile node (MN) is to search for a foreign agent (FA). For a highly
dynamic wireless network in which MNs move at great speed, even
a time interval of the order of a second between these messages
is too long. If an MN does not receive any address in response to
its solicitation messages, then to avoid network flooding, the MN
should exponentially reduce the rate of sending the solicitation
messages.

Tunnelling and encapsulation

Tunnelling establishes a virtual pipe for the packets available between
a tunnel entry and an endpoint. Tunnelling is the process of sending a
packet via a tunnel and it is achieved by a mechanism called encapsulation.
68  Fundamentals of Mobile Computing

Encapsulation refers to arranging a packet header and data in the data

part of the new packet. On the other hand, disassembling the data part of
an encapsulated packet is called decapsulation. Whenever a packet is sent
from a higher protocol layer to a lower protocol layer, the operations of
encapsulation and decapsulation usually take place.

4.2  Packet Delivery

Let us consider the situation, where the corresponding node (CN) wants to
send an IP packet to a mobile node. CN sends the packet to the IP address
of the mobile node as shown in step 1 of Fig. 4.1. The IP address of the MN
is the destination address, whereas the address of CN is the source address.
The packet is passed to the Internet that does not have any information
about the MN’s current location. So the Internet routes the packet to the
router of the MN’s home network. The home agent examines the packet to
determine whether the MN is present in its current home network or not.
In case that MN is not present, then the packet is encapsulated by a new
header that is placed in front of the existing IP header. The encapsulated
packet is tunnelled to the COA, which act as the new destination address
and the HA acts as the source address of the packet as shown in step 2 of
Fig. 4.1.
The encapsulated packet is routed to the foreign agent which
performs decapsulation to remove the additional header and forwards
the decapsulated packet to the MN, which is the actual destination, as
specified by the source node (CN), shown in step 3 of Fig. 4.1.
The MN after receiving the packet from CN, forwards a reply packet
to the CN by specifying its own IP address along with the address of the
CN as shown in step 4 of Fig. 4.1. The MN’s IP address acts as the source
address and the CN’s IP address acts as the destination address. The packet
is routed to the FA. After receiving the packet, FA forwards the packet to
CN.

4.3  Overview of Mobile IP

The goal of mobile IP is to enable packet transmission efficiently without

any packet loss and disruptions in the presence of host and/or destination
mobility.
Consider a scenario adapted from that discussed by Kozierok† to
explain the mobile IP. Suppose a person working as a business development
executive for a company needs to take care of many regional offices in India

† Kozierok, M. Charles, TCP/IP Guide: A Comprenhensive, Illustrated Internet Protocol

Reference, Starch Press, 2005.
 Mobile Internet Protocol   69

and abroad. His home office is in Delhi where he spends about 40% of his
time. The rest of the time he spends between the other offices, say, Kolkata,
Mumbai, Chennai, Kathamandu and Singapore.
A problem that arises in this context is: how does he make arrangements
so that he would continue to receive postal mails regardless of his location?
If we can answer this, we can easily understand how IP works in the
context of a mobile device.
There are two broad categories of solutions to this problem being
faced by the business executive: (i) address changing, (ii) decoupling mail
routing from his address. It would be difficult for the business development
executive to inform about his changed address to all those who are likely
to write letters to him each time he moves. Also, by the time, he would
have informed everyone about his new address, it would have become
time for the address to change again. And he certainly cannot decouple the
routing of mail from his address, unless he has set up his own personal
postal system.
A practical solution to this problem is mail forwarding. Let us say that
he leaves Delhi for Singapore for a couple of months. He will inform the
Delhi post office that he will be in Singapore. The Delhi post office would
intercept his mails headed for his normal Delhi address, relabel them, and
forward them to Singapore. Depending on where he is staying, this mail
might be redirected either straight to a new address in Singapore, or to a
Singapore post office where he can pick it up. If he leaves Singapore to go
to another city, say, Kathamandu, he would just call the Delhi post office
and tell them about his new location. When he gets back to home office,
he will cancel the forwarding arrangement and get his mail as usual. The
advantages of this system are many. It is a relatively simple mechanism
to understand and implement. A positive aspect of this scheme is that it
is transparent to everyone sending mails; they still send mail to him at his
Delhi address and the mail reaches wherever he goes. The handling of the
forwarding process is taken care by the Delhi post office as well as the
post office where he presently gets located; the rest of the postal system
does not even know that anything out of the ordinary is going on.
There are some disadvantages of this approach too. The Delhi post
office may allow occasional forwarding for free, but would probably charge
a fee if the business development executive needs this service on a regular
basis. He may also need a special arrangement in the city where he travels
to. He has to keep communicating with his home post office each time he
moves. And perhaps most importantly, every piece of mail has to be sent
through the system twice—first to Delhi and then to wherever he moves,
which is inefficient and introduces additional delay in delivering and also
loads the postal system.
Mobile IP works in a manner very similar to the postal mail forwarding
system. Each network can be considered like a different “city”, and the
internetwork of routers is like the postal system. The router that connects
any network to the Internet is like that network’s “post office”, from an
IP perspective.
70  Fundamentals of Mobile Computing

The mobile node is normally resident on its home network, which

is the one indicated by the network ID in its IP address. Devices on the
internetwork always route using this address, so the pieces of “mail”
(datagrams) always arrive at a router at the device’s “home”. When
the device “travels” to another network, the home router (“post office”)
intercepts these datagrams and forwards them to the device’s current
address. It may send the datagrams straight to the device using a new,
temporary address, or it may send them to a router on the device’s current
network (the “other post office”, Singapore) for final delivery. The mobile
node’s home router serves as the home agent and the router in Singapore
as the foreign agent. The mobile has been assigned a temporary “care-of
address” to use in Singapore (which in this case is a co-located care-of-
address). As per mobile IP terminology, the home agent tunnells the packet
to the COA.

BOX 4.4  An analogy with Mobile IP

Consider a second analogy with Mobile IP. Suppose you are moving from one
residential flat to another. What do you do?
• Leave a forwarding address with your old post office.
• The old post office forwards mail to your new post office, which then delivers
it to you.

The steps used in the operation of mobile IP are the following:

Step 1:  The remote client sends a datagram to the MN using its home
address. It reaches the home agent (say Delhi) as usual.
Step 2:  The home agent encapsulates that datagram in a new packet and
sends it to the foreign agent (say Singapore).

4.4  Desirable Features of Mobile IP

Some of the features required of mobile IP are the followings.

Transparency:  A mobile end-system should continue to keep its

IP address and there should not be any disruption of communication
after any movement. In other words, the IP address is to be managed
transparently and there should not be any effect of mobility on any ongoing
communication.

Compatibility:  Mobile IP should be compatible with the existing Internet

protocols.

Security:  Mobile IP should, as far as possible, provide users with secure

communications over the Internet.
 Mobile Internet Protocol   71

Efficiency and Scalability:  In the event of worldwide support, there can

be a large number of mobile systems in the whole Internet. This should
neither result in large number of messages nor should it incur too much
computational overhead. It should also be scalable to support billions of
moving hosts worldwide.

4.5  Key Mechanism in Mobile IP

Mobile IP is associated with the following three basic mechanisms:

• Discovering the care-of-address
• Registering the care-of-address
• Tunnelling to the care-of-address
A schematic diagram of Mobile IP is shown Fig. 4.2. The specific
protocols used by the basic mechanisms have also been shown. Observe
that the registration process works over UDP and the discovery protocol
over ICMP.

Figure 4.2  A schematic model of Mobile IP.

Discovering the care-of-address

Each mobile node uses a discovery protocol to identify the respective home
and foreign agents. The discovery of the care-of-address consists of four
important steps.
1. Mobile agents advertise their presence by periodically broadcasting
the agent advertisement messages.
2. The mobile node receiving the agent advertisement message observes
whether the message is from its own home agent and determines
whether it is on the home network or on a foreign network.
3. If a mobile node does not wish to wait for the periodic advertisement,
it can send out agent solicitation messages that will be responded to
by a mobility agent.
72  Fundamentals of Mobile Computing

The process of agent advertisements, involves the following activities:

• Foreign agents send messages to advertise the available care-of-
addresses.
• Home agents send advertisements to make themselves known.
• Mobile hosts can issue agent solicitations to actively seek information.
• If a mobile host has not heard from the foreign agent to which its
current care-of-address belongs, it takes up another care-of-address.

Registering the care-of-address

If a mobile node discovers that it is on the home network, it operates
without requiring any mobility services.
If a mobile node obtains a care-of-address from a foreign agent, then
this address should be registered with the home agent. The mobile node
sends a request for registration to its home agent along with the care-of-
address information whenever the home agent receives the registration
request information. The routing table is updated and it sends back the
registration reply to the mobile node. The mobile node makes use of the
registration procedure to intimate the care-of-address to a home agent.
These steps are schematically shown in Fig. 4.3. We now elaborate these
different steps of the registration process.

Figure 4.3  Registration process in Mobile IP.

The registration process shown in Fig. 4.3 consists of the following

steps:
1. If the mobile node is on a new network, it registers with the foreign
agent by sending a registration request message which includes the
permanent IP address of the mobile host and the IP address of its
home agent.
2. The foreign agent in turn performs the registration process on behalf
of the mobile host by sending a Registration Request containing the
 Mobile Internet Protocol   73

permanent IP address of the mobile node and the IP address of the

foreign agent to the home agent.
3. When the home agent receives the Registration Request, it updates
the mobility binding by associating the care-of-address of the mobile
node with its home address.
4. The home agent then sends an acknowledgement to the foreign agent.
5. The foreign agent in turn updates its visitors list by inserting the
entry for the mobile node and relays the reply to the mobile node.

Box 4.5  Security in Mobile IP

Security is very important in Mobile IP as mobile nodes are often connected
to the Internet via wireless links which are very vulnerable to security attacks.
For example, during the registration procedure the home agent should be
convinced that it is getting the authentic registration request from a mobile
node. Mobile IP solves this problem by specifying a security association between
the home agent and the mobile node.

Tunnelling to the care-of-address

Tunnelling takes place to forward an IP datagram from the home agent to
a care-of-address. This involves carrying out the following steps:
• When a home agent receives a packet addressed to a mobile host,
it forwards the packet to the care-of-address using IP-within-IP
(encapsulation).
• Using IP-within-IP, the home agent inserts a new IP header in front
of the IP header of any datagram.
• Destination address is set to the care-of-address.
• Source address is set to the home agent’s address.
• After stripping out the first header, IP processes the packet again.
The tunnelling operation in mobile IP and IP-within-IP encapsulation
(embedding) are shown in Fig. 4.4.

4.6  Route Optimization

In the mobile IP protocol, all the data packets to the mobile node go through
the home agent. Because of this there will be heavy traffic between HA and
CN in the network, causing latency to increase. Therefore, the following
route optimization needs to be carried out to overcome this problem.
• Enable direct notification of the corresponding host
• Direct tunnelling from the corresponding host to the mobile host
• Binding cache maintained at the corresponding host
The mobile IP scheme needs to support the four messages shown in Table 4.1.
74  Fundamentals of Mobile Computing

Figure 4.4  IP encapsulation in mobile IP.

The association of the home address with a care-of-address is called

binding.

TABLE 4.1  Messages Transmitted in Optimized Mobile IP

Message type Description

1. Binding request If a node wants to know the current location of a mobile
node (MN), it sends a request to home agent (HA).

2.  Binding On request, the node will return an acknowledgement

acknowledgement message after getting the binding update message.
3. Binding update This is a message sent by HA to CN mentioning the
correct location of MN. The message contains the fixed IP
address of the mobile node and the care-of-address. The
binding update can request for an acknowledgement.
4. Binding warning If a node decapsulates a packet for a mobile node (MN),
but it is not the current foreign agent (FA), then this node
sends a binding warning to the home agent (HA) of the
mobile node (MN).

4.7  Dynamic Host Configuration Protocol (DHCP)

DHCP was developed based on bootstrap protocol (BOOTP). DHCP

provides several types of information to a user including its IP address. To
manage dynamic configuration information and dynamic IP addresses, IETF
 Mobile Internet Protocol   75

standardized an extension to BOOTP known as dynamic host configuration

protocol (DHCP). The DHCP client and server work together to handle the
roaming status and to assign IP address on a new network efficiently. The
DHCP server allocates an IP address from a pool of IP addresses to a client.

BOX 4.6  The BOOTP protocol

The BOOTP protocol is used for booting (starting) computers from the network.
These are popularly used in case of diskless computers. Whenever a client
requests an IP address from the server machine, BOOTP searches a table which
matches to its physical address. The BOOTP protocol does not handle the
mobility-related issues since it cannot address the following problems:
•  When a host moves from one network to another network.
•  When a host seeks a temporary Internet Protocol (IP) address.

4.7.1  Significance of Dynamic Host Configuration Protocol

DHCP is an extension to the BOOTP and compatible with it. For example, if
a host is running BOOTP, it can also request configuration (example: static
configuration) from a DHCP server node. The importance of DHCP in a
mobile computing environment is that it provides temporary IP addresses
whenever a host moves from one network to another network.
DHCP supports the following three important mechanisms for IP
address allocation:

Automatic allocation:  In automatic allocation, DHCP assigns a permanent

IP address to a particular client.

Dynamic allocation:  In dynamic allocation, DHCP assigns IP address to

a client for a specific period of time.

Manual allocation:  In manual allocation, a client’s IP address is assigned

by the network administrator, where the DHCP is used to inform the
address assigned to clients.

SUMMARY

It is necessary to maintain the same IP addresses of a mobile host even

when it moves. The traditional IP design does not support mobility. So,
whenever a computer changes its location, it would need a new IP address
to be set on it according to the network address of the new network to
which it has moved in. This is inconvenient, but once given a new IP,
there would be no way of receiving the messages sent to its old IP address.
76  Fundamentals of Mobile Computing

Mobile Internet Protocol (Mobile IP) has become a standard protocol

to allow users to maintain connectivity with their home IP addresses
regardless of their physical movement. Without getting into too much
details, this chapter gives an overview of the working of mobile IP.

FURTHER READINGS

Jiannong Cao, Liang Zhang, S.K. Das, and Henry Chan, “Design and
Performance Evaluation of an Improved Mobile IP Protocol,” INFOCOM
2004.
Johnson, D., C. Perkins, and J. Arkko, “Mobility Support in IPv6,” Request
for Comments (Proposed Standard) 3775, Internet Engineering Task
Force, June 2004. Available online at http://www.rfc-editor.org/rfc/
rfc3775.txt.
Liu Yu, Ye Min-hua, and Zhang Hui-min, “The handoff schemes in mobile
IP,” Vehicular Technology Conference, 57th IEEE Semiannual Volume 1,
22–25 April 2003, pp. 485–489.
Perkins, E. Charles, “IP Mobility Support for IPv4,” Request for
Comments (Proposed Standard) 3344, Internet Engineering Task Force,
August 2002. Available online at http://www.ietf.org/rfc/rfc3344.txt.
Perkins, E. Charles, “Mobile IP,” Communications Magazine, IEEE Vol. 40,
Issue 5, May 2002, pp. 66–82.
Rajeev Koodli, “Fast Handovers for Mobile IPv6,” Internet Draft, Internet
Engineering Task Force, draft-ietf-mipshop-fast-mipv6-03.txt, October
2005. Available online at http://www.ietf.org/internet-drafts/draft-
ietf-mipshop-fast-mipv6-03.txt.
Rajeev Koodli and Charles E. Perkins, “Mobile IPv4 Fast Handovers,”
Internet Draft, Internet Engineering Task Force, draft-ietf-mip4-
fmipv4-00.txt, February 2006. Available online at http://www.ietf.
org/internet-drafts/draft-ietf-mip4-fmipv4-00.txt.
Stefan Raab and Madhavi W. Chandra, Mobile IP Technology and Applications,
Cisco Press, 2005.

EXERCISES
1. Explain why the traditional IP cannot be used in a mobile network.
What are the main differences between the traditional IP and the
mobile IP? How does mobile IP support mobile hubs?
2. Explain the limitations of IPv4 and how are they overcome by
IPv6.
 Mobile Internet Protocol   77

3. Explain the following terms associated with mobile IP:

(a) Home Address
(b) Mobile Node
(c) Foreign Agent
(d) Foreign Network
(e) Home Network
4. Write short notes on the following:
(a) Correspondent Node
(b) Care-of-Address
(c) Agent Discovery
(d) Tunnelling and Encapsulation
5. Explain the operation of mobile IP with the help of a suitable
schematic diagram and by using suitable examples.
6. What are the disadvantages of mobile IP?
7. Explain the discovery of care-of-address in the context of movement
of a mobile to a foreign network.
8. Explain the agent advertisement procedure of mobile IP.
9. What do you mean by agent solicitation? Why are agent advertisement
messages needed?
10. What do you mean by encapsulation and decapsulation in the
context of mobile IP? Explain why are these needed.
11. Give a brief account of route optimization in mobile IP.
12. What do you mean by the term binding of a mobile node?
13. What are the main functions of DHCP? Why is DHCP needed? Can
it be used when nodes are mobile? Explain your answer.
14. Explain how mobile IP is different from DHCP.
15. State some applications of DHCP.
16. Explain the significance of Dynamic Host Configuration Protocol.
Give examples of situations where it is useful.
 Mobile Transport Layer
5
I n mobile computing applications, Transmission Control Protocol (TCP)
is possibly the most popular transport layer protocol. In fact, TCP is
the de facto standard transport layer protocol for applications that require
guaranteed message delivery. TCP is a connection-oriented protocol. UDP
(User Datagram Protocol), on the other hand, is a connectionless protocol
in the TCP/IP protocol suite and does not guarantee reliable data delivery.
However, when the traditional TCP is used in mobile computing networks,
it operates in a highly inefficient and unsatisfactory manner. TCP, therefore,
needs several special adaptations to make it suitable for use in wireless
networks. This issue forms the focus of the discussions in this chapter.

BOX 5.1  A rough analogy of connectionless transmission

A rough analogy of connectionless transmission is like posting a long letter
by splitting it into several small parts, writing these parts using a number of
postcards, assigning sequence numbers to postcards, and posting these in
a post box. At the receiver-end, some postcards might be received out of
sequence, a few may arrive at the destination after undue delay, and some
postcards might even get misplaced before they arrive at the destination. On
the other hand, in a connection-oriented protocol, explicit connection is set up
between a sender and a receiver before a message is split into parts and sent
across. This connection establishment enables the receiver to know beforehand
how many postcards are expected to arrive, and based on this information,
request for any missing postcards can be issued. Thus, it can be guaranteed
that the message will be fully reconstructed at the receiver end.

In this chapter, we first provide a brief overview of the TCP/IP protocol

suite and identify the factors that render TCP operation in mobile wireless
networks unsatisfactory. Subsequently, we outline how TCP has been
extended to make it suitable to work in mobile wireless networks.
78
 Mobile Transport Layer   79

5.1  Overview of TCP/IP

The TCP/IP protocol suite was developed by DARPA in 1969 to provide

seamless communication services across an internetwork consisting of
a large number of different networks. The TCP/IP protocol suite is a
collection of a large number of protocols. The protocol suite has been
named after the two most important protocols of the protocol suite:
Transmission Control Protocol (TCP) and Internet Protocol (IP). A few
important protocols in the TCP/IP suite and the specific protocol layer
at which they operate have been depicted in Fig. 5.1. The diagram also
depicts the specific lower layer protocols that are invoked by a protocol.
As shown in Fig. 5.1, the TCP/IP protocol stack consists of four layers of
protocols. The four layers of the protocol are: Application layer, Transport
layer, Internet layer, and Network interface layer. Of the four layers,
TCP/IP does not define any specific protocol for the network interface
layer, but allows any of the standard protocols to be used at this layer.
For this reason, in Fig. 5.1 we do not show any specific protocol for the
network interface layer.

TELNET
DNS
SMTP Application
FTP Layer Message
SNMP
HTTP

Transport
TCP UDP Layer Segment

IGMP ICMP
Internet
IP Layer Packet
ARP RARP

Network Interface
Network Interface Protocols Layer Frame

Figure 5.1  TCP/IP protocol stack.

The application programmers and end-users are mainly concerned

with the application layer protocols. The application layer protocols, in
turn, make use of the services provided by the lower layer protocols.
An application layer protocol requiring to send a message to another
application (that may possibly be running on a different host either in the
same local network or in some remote network) makes use of a transport
80  Fundamentals of Mobile Computing

layer protocol and passes it with the message to be transmitted. The

specific transport layer protocol converts the message into small parts and
attaches certain information to it. The transport layer protocol first converts
a message into segments and passes these segments to the Internet layer
protocol (IP). The IP layer protocol attaches certain information to the
segments such as the destination host address to form packets. We can say
that a TCP segment is carried in one or more IP packets. The IP passes on
the packets to the network interface layer protocol, which in turn converts
them to frames by adding certain additional information to the packets
such as checksum and then transmits them on the network.
The reverse operation takes place when a frame arrives at a host.
The network interface layer protocol removes the information added by
the corresponding network interface layer protocol at the sender-end and
passes on the packet to the IP layer. The IP layer protocol at the destination
removes the information added by the IP layer at the sender’s end and
gets back the segments and passes these to the transport layer protocol.
The transport layer protocol at the receiver strips the information added
by the transport layer protocol at the sender, reconstructs the message and
sends it to the application layer.
Note that the application layer deals with messages; the transport layer
deals with segments; the internet layer deals with packets; and the data link
layer deals with frames. We will discuss the details of the operations of a
few important protocols of the TCP/IP suite in the subsequent chapters.
Over the last two decades, the Internet has seen almost exponential
growth and now the Internet applications have become ubiquitous. The
Internet-based applications are developed predominantly by using the
client-server paradigm. In a typical Internet-based application deployment
scenario, the server is an application providing certain services and a client
that typically runs on a web browser, is primarily the requester of services.
TCP has now become the de facto transport layer protocol for client-server
communications.

5.2  Terminologies of TCP/IP

In the following, we briefly discuss a few of the protocols and terminologies

associated with the TCP/IP protocol suite.

TCP (Transmission Control Protocol):  On the sending side, TCP is responsible

for breaking a message into small parts, adding sequence numbers and
certain other information and after this, making them known as segments.
TCP passes the segments to the lower layer protocol for transmission over
the network. While at the receiver’s end, TCP assembles the segments
when they arrive and reconstructs the message. TCP is a reliable protocol.
Whenever a packet is lost or corrupted during transmission, TCP detects it
 Mobile Transport Layer   81

and requests the sender for retransmission. Thus, retransmission is used as

the primary mechanism by TCP for reliable data delivery to the destination.

IP (Internet Protocol):  At the host machine of an application sending a

message, IP is responsible for constructing packets (also called datagrams)
from the segments it receives from the transport layer protocol by adding
the destination host address and then passes these on to the lower layer
protocol for transmitting. On the receiver’s side, it deconstructs the
segments and then passes these to the transport layer protocol.

HTTP (Hyper Text Transfer Protocol):  The HTTP protocol is used for
communications between a web server and the client-side application
running on a web browser.

SMTP (Simple Mail Transfer Protocol):  The SMTP protocol is used for
sending and receiving e-mails by a mail client.

MIME (Multipurpose Internet Mail Extensions):  The MIME protocol lets

the SMTP encode multimedia files such as voice, picture, and binary data
in e-mails and transmit them across TCP/IP networks. SMTP has been
designed to handle only the text contents in e-mails. MIME helps e-mails
to include non-text contents such as picture, voice, and binary data files
by encoding the binary data in the ASCII text format.

FTP (File Transfer Protocol):  The FTP protocol is used to transfer files
between the computers.

SNMP (Simple Network Management Protocol):  The SNMP protocol is used

for administration and management of computer networks. The network
manager uses tools based on this protocol to monitor network performance.

ICMP (Internet Control Message Protocol):  The ICMP protocol runs on all
hosts and routers and is mainly used for reporting errors such as a non-
reachable host.

ARP (Address Resolution Protocol):  The ARP protocol is used by IP to find

the hardware address (also called the physical address) of a computer
based on its IP address. The hardware (physical) address is stored in the
ROM (Read Only Memory) of the computer’s network interface card. It
is also known as MAC (Media Access Control) address and also as an
Ethernet hardware address (EHA).

RARP (Reverse Address Resolution Protocol):  The RARP protocol is used by

IP to find the IP address based on the physical (MAC address) address of
a computer.

BOOTP (Boot Protocol):  The BOOTP protocol is used for booting (starting)
a diskless computer over a network. Since a diskless computer does not
store the operating system program in its permanent memory, the BOOTP
82  Fundamentals of Mobile Computing

protocol helps to download and boot over a network, using the operating
system files stored on a server located in the network (discussed in
Section 4.7).

Routers:  A router is responsible for routing the packets that it receives

to their destinations based on their IP addresses, possibly via other
routers.

DNS:  It stands for Domain Name System (or Service or Server). It is a

software service available on the Internet that is responsible for translating
domain names into IP addresses. We use domain names while accessing
any website since these are alphabetic character strings that are much
easier to remember compared to the conventional IP address specification
using dot-separated numerical values. Of course, when we specify a
website (URL) using its domain name, a DNS service hosted on the Internet
translates the domain name into the corresponding IP address, since, after
all, the Internet works using IP addresses. For example, the domain name
www.iitkgp.ernet.in might get translated by the DNS to 144.16.192.245.

IP Addresses:  Each computer must have an IP address before it can

be meaningfully connected to the Internet. A packet gets routed to its
destination based on its IP address.

IGMP (Internet Group Management Protocol):  The IGMP protocol is used by

hosts to exchange information with their local routers to set up multicast
groups. A setup of multicast groups allows efficient communication,
especially for video streams and certain gaming applications. The routers
also use the IGMP to check whether the members of a known group are
active or not.

5.3  Architecture of TCP/IP

As already stated in Section 5.1, the TCP/IP protocol consists of four layers
as shown in Fig. 5.2. These layers are: Application layer, Transport layer,
Internet layer, and Network access layer. The functionalities of each of
these layers are discussed below:

Application Layer
Transport Layer
Internet Layer
Network Access Layer

Figure 5.2  TCP/IP protocol layers.

 Mobile Transport Layer   83

Application layer:  The protocols at this layer are used by applications

to establish communication with other applications which may possibly
be running on separate hosts. Examples of application layer protocols are
http, ftp, and telnet.

Transport layer:  It provides reliable end-to-end data transfer services.

The term end-to-end means that the end points of a communication link
are applications or processes. Therefore, sometimes protocols at this layer
are also referred to as host-to-host protocols. Remember that there can be
several applications or processes running on a host. Thus, to identify the
end point, it is not only the computer that needs to be identified, but also
the exact process or application that would receive the message needs
to be identified. This is efficiently accomplished by using the concept of
a port number. An application or a process specifies a port number on
which it would receive a message. Once a message reaches a host, it is
demultiplexed using the port number at the transport layer for delivery
to the appropriate application. The transport layer provides its services by
making use of the services of its lower layer protocols. This layer includes
both connection-oriented (TCP) and connectionless (UDP)  protocols.

Internet layer :  The Internet layer packs data into data packets that are
technically known as IP datagrams. Each IP datagram contains source and
destination address (also called IP address) information that is used to
forward the datagrams between hosts and across networks. The Internet
layer is also responsible for routing of IP datagrams. In a nutshell, this
layer manages addressing of packets and delivery of packets between
networks using the IP address. The main protocols included at the Internet
layer are IP (Internet Protocol), ICMP (Internet Control Message Protocol),
ARP (Address Resolution Protocol), RARP (Reverse Address Resolution
Protocol) and IGMP (Internet Group Management Protocol).

Network access layer:  The functions of this protocol layer include encoding
data and transmitting at the signalling determined by the physical layer.
It also provides error detection and packet framing functionalities. As
we will discuss later in Section 5.6, the functionalities of this layer actually
consist of the functionalities of the two lowermost layers of the ISO/
OSI protocol suite, namely data link and physical layers. The data link
layer protocols help deliver data packets by making use of physical layer
protocols. A few popular data link layer protocols are Ethernet, Token
Ring, FDDI, and X.25. Ethernet is possibly the most common data link layer
protocol. The physical layer defines how data is physically sent through
the network, including how bits are electrically or optically signalled by
hardware devices that interface with a network medium, such as coaxial
cable, optical fibre, or twisted pair of copper wires.
84  Fundamentals of Mobile Computing

5.4  An Overview of the Operation of TCP

When a client-server application runs on hosts that are wide apart,
data transmission between the client and the server may span multiple
networks. These networks are called sub-networks. For data routing, the
Internet Protocol (IP) requires that each host in the network should have
a unique address. Identification of hosts is not enough for data delivery,
the packets must be forwarded to the exact application (or to a process
in an application) requiring the packet. Within each host, every process is
identified by a port number based on which the TCP can deliver data/
information to each relevant process.

BOX 5.2  Multiplexing and demultiplexing

A host can run many client and server applications. Therefore, these different
applications can send/receive data concurrently and independently. As a result,
data sent by different applications need to be multiplexed together, before
these are sent on the network. Similarly, the TCP receives segments that may
correspond to different applications running on a host. Therefore, on receiving
a segment from its lower layer, TCP has to decide as to which application is
the recipient. This is called demultiplexing. TCP performs multiplexing and
demultiplexing by using port numbers.

Usually a message in the form of a block of data is passed to TCP

by the sending application. The TCP breaks it into many small parts and
attaches certain control information (called TCP header) to each small part.
Each small part of the data along with the TCP header is called a segment
(Fig. 5.3).

TCP Portion of
Header Data

Control
Information

Figure 5.3  The structure of a TCP segment.

The TCP header includes several items of information including the

following:
(i) Destination Port
(ii) Checksum
(iii) Sequence number
For a more detailed treatment of TCP operation, the reader is referred to
Behrouz A. Forouzan1.

1. Behrouz A. Forouzan, Data Communication and Networking, 4th ed., TMH, 2007.
 Mobile Transport Layer   85

IP datagram
An IP packet is also called a datagram. A datagram is of variable length
which can be up to 65,536 bytes. It has two fields, namely header and data.
The structure of an IP datagram is schematically shown in Fig. 5.4. In the
following, we discuss some of the important fields of an IP datagram.

Figure 5.4  IP datagram structure.

Version (Ver):  The IP version number is defined in this field, e.g. IPV4
or IPV6.
Header length (Hlen):  It defines the header length as multiples of four
bytes.
Service type:  It has bits that define the priority of the datagram itself.
Total length:  This field is allotted 16 bits to define the length of IP
datagram.
Identification:  It is mainly used to identify fragmentation that belongs to
different networks. 16 bits are allotted for this job.
Flags:  It deals with fragmentation of the data.
Fragmentation offset:  It is a pointer to the offset of the data in the original
datagram.
Time to live:  This field is used to define the total number of hops that a
datagram has to travel before discarding the operation.
Protocol:  This field has 16-bits. It defines which upper layer protocol data
is encapsulated at that time, say, for example TCP or UDP or ICMP, etc.
Header checksum:  It has a 16-bit field to check the integrity of the packets.
Source address:  It is a four byte (4  8 = 32) internet address to define
the original source.
Destination address:  It is a four byte (4  8 = 32) internet address to
identify the destination of datagram.
The IP datagrams are sent to the link layer and become ready for
transmission to the first sub-network in the path to its destination. The IP
datagrams are also called packets.
86  Fundamentals of Mobile Computing

Port address
In a client-server application, often the client and server programs are
located on different host machines. The client program usually uses a
temporary port number and the server program uses a well-known (or
permanent) port number. These port numbers are used for identification
of the application. A few well-known ports used by some of the popular
TCP/IP protocols and by user applications are given in Table 5.1.

TABLE 5.1  A Few Commonly Used Well-known Port Numbers

Protocol Port
TELNET 23
SMTP 25
RPC 111
DNS 53

Data encapsulation
When the TCP segments are handed over to Internet Protocol (IP layer), this
layer appends an IP header containing the relevant control information. A
segment after this additional control data is added, is called an IP datagram
(see Fig. 5.4). The network access layer also appends its own header and
now the segment becomes known as a frame/packet. The packet header
includes important information such as the following:
(i) Facilities requests (such as priority)
(ii) Destination sub-network address

5.5  Application Layer Protocols of TCP

The application layer protocols of the TCP/IP protocol suite are shown
pictorially in Fig. 5.5. The following are the three important application
layer protocols.

Simple Mail Transfer Protocol (SMTP):  It provides an ‘electronic mail’

function, that is used for transferring messages between different hosts.
Originally, SMTP could handle text messages only. MIME helps transmit
multimedia data within an e-mail by encoding the binary multimedia data
in the ASCII format.

File Transfer Protocol (FTP):  FTP is mainly used for transferring files
from one host to another based on a user command. FTP allows both binary
and text file transfers. Each FTP connection opens two TCP connections,
one for data transfer and the other for transfer of control commands such
as put, get, etc.
 Mobile Transport Layer   87

Figure 5.5  Application layer protocols in TCP/IP protocol suite.

TELNET:  This application layer protocol lets users use a remote log-on
facility, using which a user can log-on to a remote system. Both FTP and
TELNET make use of the TCP layer. TCP forwards these data over the
network by invoking the IP layer and the IP layer in turn invokes the link
layer protocol. A problem with this type of transmission is that it becomes
easy to sniff (secretly hear) the data by using the publicly available TCP
sniffer programs such as “TCPdump”. Due to this, at present, most users
and applications use sftp (secure ftp) and ssh (secure shell) protocols. These
protocols essentially serve to encrypt before passing the data on to the TCP
layer. These protocols also perform decryption after receiving the data.

5.6  TCP/IP versus ISO/OSI Model

It is important to realize the difference between the TCP/IP protocol suite

and the ISO/OSI model. As already mentioned, TCP/IP was developed
by DARPA (Defence Advance Research Program of the US government)
in 1969. The aim was to integrate all the computers efficiently. Thus
TCP/IP networking predates ISO/OSI which was formulated in 1976.
Naturally, TCP/IP makes no reference to ISO/OSI model and splits the
network functionalities into layers quite differently.
A rough correspondence between TCP/IP protocol layers and ISO/OSI
layers has been shown in Fig. 5.6. Observe that the Internet layer roughly
corresponds to the network layer of the ISO/OSO model. The network
access layer encompasses the data link and physical layers. The TCP/IP
protocol suite does not define specific data link layer protocols to be used
and can work on any data link protocol such as token ring and Ethernet.

5.7  Adaptation of TCP Window

The TCP primarily deploys a flow control technique to control congestion

in a network. Traffic congestion occurs when the rate at which data is
injected by a host into the network exceeds the rate at which data can be
delivered to the network. A flow control technique helps adapt the rate
88  Fundamentals of Mobile Computing

Figure 5.6  A comparison of TCP/IP and ISO/OSI models.

of data transmission by the TCP at the sending host end. The flow control
technique helps to prevent the build-up of congestion in the network and
at the same time helps to prevent buffer overrun at the slow receivers.
If data transmissions occur at a much faster rate than what the network
infrastructure can comfortably support, then data packets get built up at
the routers. When the buffers at routers start to overflow, the packets start
getting lost. Additionally, if data transmissions by a sender take place
at a much faster rate than what a slower receiver can handle, then the
receiver’s buffer starts to get flooded and hence the packets get lost. TCP
handles both these causes of packet loss by reducing the rate at which
data is transmitted at the sender’s end. Thus, a receiver uses the flow
control mechanism to restrict how fast a sender can transmit. However,
to provide an acceptably fast data transmission service, once congestion
disappears the transmission rate at the sender’s end needs to be increased
to a suitable value. Thus we can say that a flow control technique helps
TCP dynamically adjust the transmission rate at the sender’s end, reducing
the transmission rate as congestion starts to develop and increasing it
as congestion starts to disappear. The flow control mechanism deployed
by TCP (called the sliding window protocol) is primarily based on the
concepts of congestion window and advertised window. We now briefly
describe these techniques.
When a sender starts to send data packets, the receiver indicates an
advertised window (or receiver window) to the sender while sending
acknowledgements. The advertised window is usually set equal to the size
of the receive buffer at the receiver. The sender uses the advertised window
size obtained from the receiver to determine the maximum amount of data
 Mobile Transport Layer   89

that it can transmit to the receiver without causing buffer overflow at the
receiver. In other words, to prevent buffer overflow at the receiver, the data
packets transmitted by a sender without having received acknowledgments
for them should not exceed the size of the buffer available at the receiving
end.
For each segment sent, a sending host expects to receive an acknow-
ledgment. A congestion window indicates the maximum number of
segments that can be outstanding without the receipt of the corresponding
acknowledgement before the TCP at the sender’s end pauses transmitting
and waits for an acknowledgement to arrive. The TCP at a sender’s end
pauses if the number of segments for which the acknowledgement is
outstanding becomes equal to the congestion window. A sender sets the
congestion window size to 1 and keeps on increasing it until duplicate
acknowledgements are received or until the number of outstanding packets
becomes equal to the size of the advertised window.
Upon receipt of an acknowledgement, TCP detects packet loss using
Retransmission timeout (RTO) and duplicate acknowledgements. After
transmitting a segment, a TCP sender sets the retransmission timer for only
one packet. If an acknowledgement for the packet is not received before
the timer goes off, the packet is assumed to be lost. RTO is dynamically
calculated. Timeouts can take too long. Again the TCP sender assumes that
a packet loss has occurred if it receives three duplicate acknowledgements
consecutively. In TCP, when a receiver does not get a packet that it expects
in a sequence but gets an out of order packet, it considers that the expected
packet might have got lost and it indicates this to the sender by transmitting
an acknowledgment for the last packet that was received in order. Thus,
three duplicate acknowledgement are also generated if a packet is delivered
at least three places beyond its in-sequence location.
In wired networks, packet losses primarily occur on account of
congestions encountered in the transmission path. However, in a wireless
environment packet losses can also occur due to mobility and channel
errors. In wired networks, bit errors are rare. On the other hand, wireless
networks are vulnerable to noise. Noise can cause intermittent bit errors.
Further, there can be intermittent disconnections due to fading and also
due to obstructions that may be encountered by a mobile host. Further,
packets may get lost during handoff. An intermittent disconnection may
cause the TCP at the sender’s end to time out for an acknowledgement
and cause it to retransmit. This would cause unnecessary retransmissions
to occur, even though the packet may be buffered at a router. Also, various
additional causes of packet loss can result in a high rate of packet loss in
a mobile wireless network. These losses would be interpreted by TCP as
symptoms of congestion and force it to operate in slow start. This would
cause the network to operate inefficiently and result in unacceptable slow
data transmission.
90  Fundamentals of Mobile Computing

BOX 5.3  Adaptive transmission control mechanism

When many packets are transmitted to a single receiver and the rate with
which these packets are transmitted is higher than the processing rate of the
destination host or an intermediate router, the buffers of the router (or the
destination as the case may be) get filled quickly. This results in dropping packets
at the affected router or the destination. When a sender realizes that some
packets might have been dropped based on acknowledgements not arriving
before timeout or based on the receipt of three duplicate acknowledgements,
possibly due to congestion, it tries to retransmit the missed packets. Though
this mechanism may overcome packet losses, it does not resolve the congestion
problem. The congestion problem is resolved through an adaptive transmission
control mechanism called flow control.

5.8  Improvement in TCP Performance

TCP was designed for traditional wired networks. If used as it is, a few
shortcomings become noticeable. We first review a few relevant aspects
of traditional TCP. Then we discuss how TCP has been extended to work
efficiently in a mobile environment.

5.8.1  Traditional Networks

In the wired networks, packet losses are primarily attributable to congestions

that get built-up in the network. To reduce congestion, TCP invokes the
congestion control mechanisms. Congestion control is primarily achieved
by reducing the transmission window, which in turn results in slower data
transfer. The important mechanisms used by TCP for improving (tcp-reno
model) performance are given below.

Slow start
The slow-start mechanism is used when a TCP session is started.
Instead of starting transmission at a fixed transmission window size,
the transmission is started at the lowest window size and then doubled
after each successful transmission. The rate of doubling is tied to the rate
at which acknowledgements come back. Thus, the doubling of window
size occurs at every round trip time (RTT). RTT is the time that elapses
between a segment is transmitted by a sender and the corresponding
acknowledgement is received. If congestion is detected (indicated by
duplicate acknowledgements), the transmission window size is reduced to
half of its current size and the congestion avoidance process starts. We can
say that this mechanism of rate doubling and reduction to half the previous
value is nothing but a binary search technique deployed to determine the
‘right’ transmission window size.
 Mobile Transport Layer   91

The slow-start process begins by the sender setting the transmission

window size to one, i.e., transmiting one segment to the receiver. The sender
does not transmit the next segment until it receives an acknowledgement
for the previous segment. Once the acknowledgement is received, the
sender becomes sure that the congestion window (network capacity) is at
least one segment. To determine the exact congestion window size, the
sender doubles the transmission window size. It transmits two segments
and after arrival of the two corresponding acknowledgements, it again
increases the transmission window size by two and sets it equal to four,
and so on. Increments that occur to the size of the congestion window
are, thus, exponential. A congestion window is doubled every time the
acknowledgements arrive smoothly. This exponential growth of congestion
window stops at the congestion threshold.

Congestion avoidance
The congestion avoidance algorithm starts where the slow start stops.
Once the congestion window reaches the congestion threshold level, then
after that if an acknowledgement is received the window size is increased
linearly, i.e., the window size doubling is avoided. From this point, the TCP
increases its transmission rate linearly by adding one additional packet to
its window at each transmission time. If congestion is detected at any point,
the TCP reduces its transmission rate to half the previous value. Thus the
TCP seesaws its way to the right transmission rate. Clearly, this scheme is
less aggressive than the slow-start phase (note the linear increase against
the exponential growth in slow start).

Fast retransmit/fast recovery

Usually, a sender initiates a timer after transmitting a packet and sets the
timeout value (RTO). RTO is calculated based on RTT. The sender waits
for an acknowledgement of a transmitted packet from the receiver until the
timer expires. When the timer expires, it retransmits the packet. However,
there exists another situation under which a sender can retransmit a packet.
This mechanism is called fast retransmission. In this, the retransmission is not
triggered by a timer, it is rather triggered by the receipt of three duplicate
copies of an acknowledgement for a packet received from the sender. Since
duplicate acknowledgements also arise when a segment is received out of
order, the sender waits for three copies of acknowledgements for the same
packet. This is taken by the sender as the confirmed indication of a missed
packet for starting to retransmit the particular packet. When retransmission
occurs, the congestion window size is reduced by half. For example, if
the current congestion window size is four segments, then it is set to two
segments. Once the lost segment has been retransmitted, TCP tries to
maintain the current data transmission rate by not going back to slow start.
This is called fast recovery. In fast recovery, the congestion window size
is incremented by three since the retransmission occurred after the third
92  Fundamentals of Mobile Computing

duplicate acknowledgement. This is construed to be the indication that

three packets would have been successfully buffered at the receiver end.
Thus, in fast recovery, compensation for the segments that have already
been received by the receiver is carried out. If the acknowledgements are
received smoothly, it is considered to be the indication that there is no
congestion.

5.8.2  TCP in Mobile Networks

In Internet, TCP is the de facto standard transport protocol. It has been

remarkably successful in supporting the diverse applications which drive the
Internet’s popularity. A few of such applications are access to information
hosted across various websites, file transfer and email. The performance
of TCP is considered satisfactory for wired networks, but it suffers from
serious performance degradation in wireless networks. Wired networks are
significantly different from wireless mobile networks. The main differences
are much lower bandwidth, bandwidth fluctuations with time and also as a
mobile host moves, higher delay, intermittent disconnections, high bit error
rate, and poor link reliability. An implication of these differences is that
packet losses are no longer only due to network congestion, they may well
be also due to intermittent link failures, bit errors due to noise, or handoffs
between two cells. Therefore, the traditional TCP assumptions are not
valid in mobile (wireless) environments. This leads to poor performance
of TCP in mixed wired-wireless environments. Several modifications at the
transport layer have been proposed and studied in recent years to deal with
the problems. To understand these works, we need to first consider single-
hop wireless networks (such as wireless LANs). Based on this, we shall
discuss multi-hop wireless networks (such as mobile ad hoc networks). A
few important mechanisms used to improve TCP performance over mobile
wireless networks are discussed below.

5.8.3  TCP in Single-hop Wireless Networks

We first discuss the modifications proposed to TCP to make it effective in

single-hop wireless networks.

Indirect TCP (I-TCP)

This protocol was proposed by Bakre and Badrinath2. It segments the
connection between the fixed host and the mobile host into two different
connections: the wired part and the wireless part (Fig. 5.7). The wired
connection exists between the fixed host (FH) and the base station (BS)
and the wireless part connection exists between the BS and the mobile

2.  Bakre, A. and B.R. Badrinath, “I-TCP: Indirect TCP for Mobile Hosts,” Proceedings
15th International Conference on Distributed Computing Systems (ICDCS), May 1995.
 Mobile Transport Layer   93

Figure 5.7  A schematic of working of indirect-TCP.

host (MH). Thus, the base station maintains two separate TCP connections:
one over the fixed network and the other over the wireless link. The
wireless link usually has poor quality communication, but it gets hidden
from the fixed network at the BS. When a packet is sent by FH to MH, the
packet is received by BS first and then BS transmits it to the MH over the
wireless link. If the mobile host moves out of the current BS region, the
whole connection information and responsibilities that are with the current
BS are transferred to the new BS.
The advantage of the split connection approach of I-TCP is that it
does not need any changes to be made to the standard TCP protocol.
By partitioning a TCP connection into two connections in I-TCP, the
transmission errors in the wireless part would not propagate into the fixed
network, thereby effectively achieving an increase in bandwidth over the
fixed network. An important disadvantage of this scheme is that I-TCP does
not maintain the semantics of TCP as the FH gets the acknowledgement
before the packet is delivered at MH. I-TCP does not maintain the end-to-
end semantics of TCP and assumes that the application layer would ensure
reliability.

Fast retransmission
This approach was suggested by Caceres et al.3 to overcome the delay in
transmissions caused due to intermittent disconnections such as those that
occurs when a mobile host (MH) moves to a foreign agent (FA) during a
TCP communication. As already discussed, TCP transmission behaviour
after a disruption depends on its duration. The extremely short disruptions
(lasting for a time much less than RTO) would appears as short bursts of

3.  Caceres, R. and L. Iftode, “Improving the Performance of Reliable Transport

Protocols in Mobile Computing Environments,” IEEE JSAC, Vol. 19, No. 7, July 2001.
94  Fundamentals of Mobile Computing

packet losses. In this case, the TCP retransmits those packets for which the
timeout occurs and recovers them without slow-start. However, for long
disruptions (lasting for a time much greater than RTO), TCP resorts to
slow-start. This results in inefficiency. As soon as a mobile host registers
at a foreign agent, it starts sending duplicate acknowledgements. As is
standard with TCP, three consecutive acknowledgements for the same
TCP segment are inferred as a packet loss by the host-end TCP, and it is
also inferred that the connection is live, thereby causing the fast retransmit
behaviour of TCP.
The advantage of this scheme is that it reduces the time for the MH
to get reconnected, otherwise FH would wait for RTO unnecessarily. The
disadvantage of this approach is that it does not propose a general approach
for TCP communication in mobile wireless networks. For example, it does
not address the specific error characteristics of the wireless medium.

Snooping TCP (S-TCP)

Balkrishnan et al.4 proposed a protocol that improves TCP performance
by modifying the software at the base station while preserving the end-to-
end TCP semantic. The modified software at the base station is known as
snoop. It monitors every packet that passes through the TCP connection in
both directions, that is from MH to FH and vice versa. It buffers the TCP
segments close to the MH. When congestion is detected during sending of
packets from the FH to MH in the form of a duplicate acknowledgement
or the timeout, it locally retransmits the packets to MH if it has buffered
the packet and hides the duplicate acknowledgement.
An advantage of snooping TCP is that it maintains the TCP semantics
by hiding the duplicate acknowledgements for the lost TCP segment and
re-sends the packets locally. However, it also suffers from higher overheads
incurred when MH moves from its current BS to a new BS, the packet
buffered at the current BS need not be transferred to the new BS.

Mobile TCP (M-TCP)

This protocol for mobile cellular networks was proposed by Kevin
Brown et al.5 In mobile wireless networks, users would badly suffer from
unacceptable delays in TCP communications and frequent disconnections
caused by events such as signal fades, lack of bandwidth, handoff, unless
these are explicitly handled by the protocol. The M-TCP protocol tries to
avoid the sender window from shrinking or reverting to slow-start when
bit errors cause a packet loss, as is attempted in I-TCP and snooping TCP.

4.  Balakrishnan, H., V. Padmanabhan, S. Seshan, and R. Katz, “A Comparison of

Mechanisms for Improving TCP Performance over Wireless Links, ”Proceedings of
ACM SIGCOMM’96, August 1996.
5.  Brown, K. and S. Singh, “M-TCP: TCP for Mobile Cellular Networks,” ACM
Computer Communication Review, Vol. 27, No. 5, October 1997.
 Mobile Transport Layer   95

In this protocol, as in I-TCP, the TCP connection between the fixed host
and the mobile host is segmented into wired and wireless parts—the wired
part connection between the fixed host (FH) and the supervisory host (SH)
and the wireless part connection between the SH and the mobile host (MH).
Many MHs are connected to SH through several base stations as shown
in Fig. 5.8. The SH supervises all the packets transmitted to MH and the
acknowledgements sent by MH. It is also used as an interface between FH
and MH and vice versa.

Figure 5.8  A schematic of operation of the M-TCP protocol.

When a packet is sent to FH by MH using SH, the wired part uses

the normal unmodified TCP and the wireless part uses the modified
version of TCP known as M-TCP to deliver data to MH. This packet is
acknowledged only when the MH receives the packet. Thus, it maintains
the TCP semantics, unlike the I-TCP. In case the acknowledgement is not
received by FH, SH decides that MH is disconnected and sets the sender
FH window size to zero. This prevents re-transmission. When SH notices
that the MH is connected, it sets the full windows size of the sender FH.
When MH moves from its current SH region to a new SH region, a state
transfer take places, so that the new SH can maintain TCP connection
between FH and MH.

Freeze-TCP
The basic idea in this scheme is to “freeze” the TCP senders’ streams, little
before a disconnection is to occur. This is done by artificially sending a “Zero
Windows Advertisement” informing the sender that the receiver cannot
96  Fundamentals of Mobile Computing

receive data at the moment. When the sender resumes its connectivity, the
receiver can unfreeze the sender by sending the value of its actual receive
window. The most interesting advantage of this proposal is the avoidance
of the slow-start period upon re-establishment of connectivity. The
freeze-TCP method has the advantage that it does not require the
involvement of the intermediate nodes and hence it can be used if the IP
payload is encrypted. This method offers a promising approach for use in
Virtual Private Networks (VPNs).

5.8.4  TCP in Multi-hop Wireless Networks

The TCP-F (TCP feedback) protocol has been proposed for extending TCP
to multiple-hop networks. In a mobile ad hoc network, a sender MH sends
a packet to destination MH through the intermediate MH, since all the
nodes of networks are MH. Again the MHs are free to move arbitrarily
which changes the network topology unpredictably. If the normal TCP runs
over in this network, there may be significant performance degradation
at the transport layer. This happens because the normal TCP is unable
to distinguish between packet loss resulting from link failure and packet
loss due to congestion on the network. As a result, the normal TCP
invokes the congestion control mechanism, even if a packet loss occurs
due to link failure. When this happens, TCP waits for a longer time to
retransmit the lost packet and this slows down the rate of transmission.
The TCP-F proposed by Chandran et al.6 addresses the problem caused by
link failure due to mobility by performing a freezing action and limiting
the re-transmissions. For simplicity, consider that a source MH is sending
packets to a destination MH. As soon as an intermediate MH detects the
disruption of route due to mobility of MH along that route, it sends a
route failure notification (RFN) packet to the source MH and records that
event. Each intermediate MH that receives the RFN packet invalidates
the particular route and prevents the incoming packets intended for the
destination MH passing through that route. If the intermediate node MH
knows of an alternate route to destination MH, this alternative route can
now be used to support further communication and the RFN is discarded.
Otherwise, the intermediate MH propagates the RFN towards the source
MH. On receiving the RFN, the source MH completely stops sending
further packets (new or retransmission), then it marks all its existing timers
as invalid and freezes the send windows. The source MH remains in
this state until it is notified of the restoration of the route through route
re-establishment notification (RRN) packets.
A summary of the proposed enhancements to the TCP to support host
mobility is presented in Table 5.2.

6.  Chandran, K., S. Raghunathan, S. Venkatesan, and R. Prakash, “A Feedback-

based Scheme for Improving TCP Performance in Ad Hoc Wireless Networks,”
IEEE Personal Communications, 8(1): 34–39, February 2001.
 Mobile Transport Layer   97

TABLE 5.2  A Comparative Study of a few Important Protocols for Mobile

Applications

TCP approach Mechanism used Merits Demerits

Indirect TCP Segments the TCP • Simple • Loss of the TCP

(I-TCP) connection into two • Isolation of wire semantics
and wireless links • Security
is possible problem
Snooping TCP Snooping of data • Transparency • Inadequate
(S-TCP) and acknowledge- • MCA interaction isolation of the
ments wireless links
• Security problem
Mobile TCP The segmented • End-to-end • Poor isolation
(M-TCP) TCP connection can segment is wireless links
choke the sender maintained • Security
through window • Handles frequent Problem
sizes disconnections
Fast retrans- It avoids slow-start • Simple • Not transparent
mission after any roaming • More efficient • Mixed layers
Fast recovery
Freeze-TCP It freezes the TCP, • Works even when • Changes in TCP
later it resumes the there are long • MAC dependent
TCP after recon- interruptions
nection

SUMMARY

In this chapter, we first reviewed some basic aspects of the TCP/IP protocol
suite as the de facto protocol for the Internet and for a host of other
applications. We then identified the important problems that might arise
when TCP is used as it is in mobile wireless networks. We also discussed
the adaptations that have been extended to make TCP work satisfactorily
in the mobile environment.

FURTHER READINGS

Ahuja, A., S. Agarwal, J.P. Singh, and R. Shorey, “Performance of TCP

over Different Routing Protocols in Mobile Ad Hoc Networks,” IEEE
Vehicular Technology Conference, Vol. 3, pp. 2315–2319, Tokyo, 2000.
Anantharaman, V., S.J. Park, K. Sundaresan, and R. Sivakumar, “TCP
Performance over Mobile Ad Hoc Networks: A Quantitative Study,”
98  Fundamentals of Mobile Computing

‘Was to appear in Wireless Communications and Mobile Computing Journal

(WCMC), Special Issue on Performance Evaluation of Wireless Networks,
2003.
Bhagwat, P., P. Bhattacharya, A. Krishna, and S.K. Tripathi, “Enhancing
Throughput over Wireless LANs Using Channel State Dependent
Packet Scheduling,” IEEE INFOCOM’96, San Francisco, March 1996.
Dyer, T.D., and R.V. Boppana, “A Comparison of TCP Performance over
Three Routing Protocols for Mobile Ad Hoc Networks,” ACM Mobihoc,
October 2001.
Fu, Z., P. Zerfos, H. Luo, S. Lu, L. Zhang, and M. Gerla, “The Impact
of Multihop Wireless Channel on TCP Throughput and Loss”, IEEE
INFOCOM’03, San Francisco , March 2003.
Holland, G. and N.H. Vaidya, “Analysis of TCP Performance over Mobile
Ad Hoc Networks,” MOBICOM’99, Seattle, August 1999.
Xiang Chen, Hongqiang Zhai, Jianfeng Wang, and Yuguang Fang, “A
Survey on Improving TCP Performance over Wireless Networks,”
Dept. of Electrical and Computer Engineering, University of Florida,
Gainesville, FL 32611, http://winet.ece.ufl.edu/~zhq/book05chen1.
pdf.

EXERCISES
1. State true or false against each of the following statements. Briefly
explain your answer in each case.
(i) TCP is a peer-to-peer, connection-oriented protocol.
(ii) The link layer is responsible to deliver data packets by making
use of lower layer protocols and also provides error detection
and packet framing functionalities.
(iii) TCP guarantees that data will be delivered without loss,
duplication or transmission errors.
(iv) TELNET encodes multimedia data using the MIME protocol.
2. Explain the following terms associated with TCP/IP stack:
(a) IP, (b) HTTP, (c) SMTP, (d) MIME, (e) FTP, (f) SNMP, (g) ICMP,
(h) ARP, (i) RARP, (j) DNS, (k) IP Addresses, (l) IGMP
3. Write short notes on:
(a) TELNET, (b) FTP, (c) SMTP, (d) TCP/IP vs. ISO/OSI protocol
model.
4. Explain the layered architecture of the TCP/IP protocol suite and
compare it with the ISO/OSI architecture.
 Mobile Transport Layer   99

5. Answer the following with respect to missing and duplicate segments

in TCP operation.
(a) What can cause segments to be missed at the receiver-end and
also cause duplicate segments to arise? Explain your answer
using a suitable scenario of operation.
(b) How exactly is a missing segment detected in TCP? Explain
the specific actions that take place when a missing segment is
detected.
6. What is slow-start in TCP operation? Explain its working. How does
slow-start help improve the performance of TCP?
7. What problems would occur if the traditional TCP is used in mobile
wireless environments? Discuss how TCP can be adapted to work
efficiently in a mobile network environment.
8. Explain Indirect-TCP (I-TCP) with the help of a suitable schematic
diagram.
9. Why are the I-TCP acknowledgements and semantics not end-to-
end? What are the implications of this?
10. What is the snooping TCP approach in mobile wireless networks?
Discuss its advantages.
11. How are handoffs handled in snooping TCP?
12. Briefly discuss the M-TCP approach of extending TCP to work
efficiently in mobile wireless networks. How does M-TCP maintain
end-to-end semantics?
13. Explain the working of Freeze-TCP. What are the shortcomings of
this protocol?
14. Why do congestions occur in a network? Explain how does TCP
detect and handle congestions.
 Mobile Databases
6
R ecent developments in technologies have made it possible for a mobile
user to have uninterrupted information and service access at his
disposal. For example, a user can even browse Internet or carry out online
banking transactions while travelling in a vehicle. This has been made
possible, among other things, by the mobile database technology that allows
the applications for hand-held devices to access data stores even while on
the move. The mobile database technology makes data from database
applications available ubiquitously to a mobile worker. An employee using
a hand-held device can link to his corporate network, download data, work
offline, and then connect to the application seamlessly. Consider another
example. Using the mobile database technology, a courier delivery worker
can collect signatures after each delivery and send the delivery information
to a corporate database as soon as the delivery is made.
Traditional database systems are difficult to use in a mobile computing
environment. Traditional databases make implicit assumptions that do
not hold in a mobile computing environment. To realize effective mobile
databases, the special characteristics of mobile computing should be
considered in the database design. For example, mobile transactions have
to cope with unexpected disconnections caused due to mobility. The
transaction processing system also needs to cope with the severely restricted
resources of the mobile hosts, including low battery life, slow processor
speed, and limited memory. Mobile applications are required to react to
frequent changes in the environment such as changing locations, high
variability of network bandwidth, and dynamically changing data. Taking
all these limitations and constraints into consideration, the different issues
in database design such as data dissemination, data replication, transaction
models, query processing and concurrency need to be investigated to
determine the specific changes that are required. Before discussing these
issues, we need to keep the basic structure of a mobile application in mind.
Figure 6.1 presents a schematic representation of invocation of the services
100
 Mobile Databases   101

of an application in a simple mobile computing system. As can be seen in

this figure, a mobile host (MH) in a wireless cell is connected to a Mobile
Support Station (MSS) that has a wireless interface. The hosts on the fixed
network do not have a wireless interface. The MSS is connected to the
fixed network though a wired fixed network. A mobile client can invoke
a service of an application server on the fixed network. The application
server, in turn, accesses the database servers using a connectivity protocol.
A few specific requirements identified for a mobile database system
are enumerated below1.
• A mobile database should have a small footprint to fit into a resource-
constrained mobile device.
• It should be able to run without the services of a database
administrator.
• It should easily interoperate with the large enterprise databases.
• A mobile database needs to support rather simple insert, delete,
update, and query functionalities.

Figure 6.1  A schematic of service invocation in a mobile computing environment.

The commercially available mobile database systems allow operations

across a wide variety of platforms and data sources. They also allow users
with a hand-held device to synchronize with the Open Database Connectivity
(ODBC) database content, and personal information management data and
e-mail from Lotus Development’s Notes or Microsoft’s Exchange. These
database technologies usually support query-by-example (QBE) or SQL

1.  Lu, E. and Y. Cheng, “Design and Implementation of a Mobile Database for
Java Phones,” Computer Standards and Interfaces, Elsevier, Vol. 26, pp. 401–410, 2004.
102  Fundamentals of Mobile Computing

instructions. Some of the commercially available Mobile relational database

systems includes IBM’s DB2 Everywhere 1.0, Oracle Lite, and Sybase’s SQL
Anywhere.
A few important factors that need to be considered for service
invocation in a mobile computing environment are as follows:
• User time is highly valuable.
• Connection time requirement should be minimum.
• The number of bytes, or packets, transferred in the unit of charges
is computed in digital cellular systems and therefore needs to be
minimized.
• Time-of-day based charges may vary based on whether communi-
cation occurs during peak or off-peak periods.
• Energy of the mobile handsets is limited. Battery power is a scarce
resource and should be optimized.

6.1  Issues in Transaction Processing

Database applications are normally structured into transactions. A transaction

is a unit of operation that changes a database from one consistent state into
another consistent state. Normally, different transactions on a database
operate in an interleaved manner. Interleaving the access of different
transactions to the database can remarkably improve the throughput and
resource utilization of the database. Therefore, one important aim in the
design of database systems is to maximize the number of transactions that
can be active at a time.
Let us now examine how the ACID properties (Atomicity, Consistency,
Isolation, Durability) are ensured in a database in the presence of interleaved
execution of transactions. While each transaction preserves the consistency
of the database at its boundaries, a transaction that fails to complete might
cause the integrity of the database to be violated. Rollback protocols are
used to ensure atomicity and durability properties when a transaction fails to
complete. Isolation can be ensured through the use of locking and rollback
protocols.
A particular sequencing of actions of different transactions is called a
database schedule. However, concurrent execution of transactions can lead
to some database schedules that take the database to an inconsistent state.
A database is in a consistent state if it satisfies all the defined semantic
integrity constraints. A transaction model defines the framework for the
definition and execution of transactions. The integrity constraints on a
database are given by the well-known ACID properties, meaning that
a transaction should be an atomic, consistent and recoverable unit and
should not interfere with other transactions that are executed concurrently.
In some application scenarios such as those in a mobile environment, the
 Mobile Databases   103

BOX 6.1  Issues in transaction processing

A transaction consists of many read and write operations on the database. In a
concurrent transaction processing environment, arbitrary transaction execution
may lead to violation of database integrity. To exclude the violation of integrity
constraints, the concept of transaction serializability has been introduced. The
execution of a set of concurrent transactions is said to be serializable, when
the concurrent execution of the transactions is equivalent to some serial
execution of these transactions. To successfully complete a transaction, all
operations must be performed on the database in some serializable order
and a recovery protocol must be executed. But, how does a DBMS ensure
serializability? In order to ensure serializability, a DBMS needs to enforce the
following constraints.
Atomicity:  Either all operations of a transaction are reflected on the database
or none at all. That is, all the operations of a transaction are together treated
as a single indivisible unit.
Consistency:  A transaction should transform the database from one consistent
state to another consistent state.
Isolation:  The effect of a transaction should be such that it appears to be
executed in isolation. That is, its intermediate results must not be seen by other
transactions that commit before it.
Durability:  Changes made to the database by a committed transaction must
persist even after any database failure, rollback, etc. that may take place.

ACID properties turn out to be too stringent for realizing any meaningful
database-based application in practice and therefore these properties
require to be relaxed.

6.2  Transaction Processing Environment

In this section, we discuss the different transaction processing environments.

6.2.1  Centralized Environment

This is the oldest data-processing system in which a large program

manages both the tasks of initiating a transaction by interacting with
users and processing it. In this environment, the overheads associated
with the processing of a transaction are rather low since only a single user
executes the transactions. Recovery is also a trivial issue here, since rollback
recoveries do not arise.
104  Fundamentals of Mobile Computing

6.2.2  Client-server Environment

In this environment, the execution of a transaction and transaction initiations

are done by a server and a client respectively. In this environment, the
server is typically concurrent. That is, many clients can submit transactions
simultaneously.

6.2.3  Distributed Environment

In this environment, data is distributed over a network. That is, either a

transaction can get fully executed at any arbitrary node, or different parts
of transaction can be executed at different nodes. Atomicity and durability
of a transaction become nontrivial issues in this environment.

6.2.4  Mobile Environment

This is a special type of distributed environment and can accommodate

user movements while issuing the transactions, and the system reports
the results. Even though it as an extension of the distributed model, it
becomes important to satisfactorily address many nontrivial issues that are
not present in a typical distributed computing scenario. These issues are
discussed in the following:

Movement of users
The movement of users introduces several complications as far as the
database operations are concerned. The complications arising due to user
mobility must be satisfactorily addressed. Users tend to move from one
cell to another while maintaining the network connection. To support this,
issues such as tracking users, and maintaining the network connection while
crossing the cell boundaries must be addressed. Recovery of transactions is
complicated as it becomes a tricky issue to determine the location at which
the logs are to be maintained. The options available are:
Maintaining logs at the mobile host:  This may be difficult due to lack
of sufficient resources such as memory and processing power. As a result,
storing the logs at the mobile hosts and initiating the recovery in case of
failure would be an unacceptable overhead. Also, as the failure rate of a
mobile host is high, this approach gets further complicated and hence not
considered satisfactory.
Maintaining logs at mobile host’s base MSS:  This approach incurs
commu-nication overhead because each transaction log must be updated,
introducing one additional round trip delay, which is undesirable.
 Mobile Databases   105

Maintaining logs at the current MSS:  In this approach, logs are maintained
at the current MSS for all transactions submitted from its cell. When an
MH moves out of the cell, all logs are transferred to the new MSS. This
approach generates a very high network traffic due to the transfer of logs.
Also, it introduces additional complications such as how much resources
should be allotted to a mobile host, particularly the foreign ones, and what
to do with the logs if an MH fails.

Disconnections
A fairly complicated issue in the operation of a mobile database arises due
to the temporary disconnections of mobile users arising out of attempts
to save power. In a typical distributed environment, the non-availability
of a host is considered to be a failure. But here one must distinguish
between a failure and a planned disconnection. In a planned disconnection,
depending on the application, the mobile user can perform some operations
by downloading data beforehand in anticipation. This is popularly called
data hoarding. Another way to deal with the disconnections is by migrating
transaction processing to a non-mobile computer.
One more approach to handle the non-availability of an MH for a
temporary period is by maintaining proxy agents at the MSS. A proxy agent
represents the MH during its absence and participates in communication and
finally hands over control to MH on its reappearance. In this approach, it is
necessary to satisfactorily take care of creation of proxy agents, participate
in communication, hand over control and ensure self-destruction.
But when an MH disappears for a long duration, the proxy agent
does not know what to do. Maintaining it in the wait state may lead to
unnecessary wastage of resources of the MSS. Collecting messages on
behalf of an MH in its absence or use of mailboxes may be an appropriate
solution. When a mobile host is not available in the network, all messages
destined for it can be routed to its mailbox, which will be retrieved by the
corresponding MH after reconnection. For implementation of this concept,
a base MSS must be present for each MH, which again is a very difficult
proposition.

Poor communication media

The bandwidth allocated to mobile users could be very low. Even when
sufficient bandwidth is available, many kinds of noise such as interference
from other traffic, atmospheric conditions, etc. may corrupt the data, thus
causing multiple retransmissions. Transmission over a wireless medium
consumes more power than that over a wired link and power is a scarce
resource for mobile hosts. For this reason, mobile hosts tend to disconnect
from the network whenever there is no data to send or receive in the near
future. This phenomenon is known as dozing.
106  Fundamentals of Mobile Computing

BOX 6.2  Energy consumption in wireless communication

A wire serves as a guided medium, whereas in wireless transmission power
is radiated in all directions. This causes the radiated energy density to fall off
exponentially with distance. Also, the impedance of air is much higher than that
of a wire. These are the two important reasons for higher energy consumption
in wireless communication compared to that in wired communication.

Processing power
Mobile hosts are generally equipped with less powerful CPUs in order to
conserve power. Due to this reason, compute intensive applications such
as a database server are difficult to use. This constraint also forces one to
rethink the way computations are done. For example, it may be required
to find ways to reduce the client-side computations.

Memory
Mobile hosts usually have a very limited main memory as well as secondary
storage. Nonvolatile memory devices such as hard disks consume significant
power for running the motor that drives the read/write heads. Also, hard
disks are usually bulky and heavy. Therefore, all the important applications
and system software requiring storage on permanent memory may be put
into a flash memory, which consumes much less power.

Battery power
Battery power is possibly the most precious and scarce resource of a
mobile host. Not only batteries store very limited amounts of energy, often
recharging them becomes difficult. Most of the battery energy is spent on
non-computational tasks such as transmitting and receiving data over the
wireless medium and some energy is also consumed by the secondary
storage and the input and output devices.

User interfaces
The I/O devices consume significant power and often are bulky. One needs
to find new ways to optimize the power they consume and also the space
they occupy. As a substitute for keyboard, one can use pen-based input
devices, which occupy less space but consume more computing power for
lengthy inputs, particularly for SQL type statements. Thus, a graphical/
symbolic interface with pointing capabilities may suit this environment.

Security
Since mobile handsets are carried across several networks, the prevention
of impersonation of one machine by another is problematic. For example,
 Mobile Databases   107

when a mobile computer enters a new cell, the specific data it sends and
receives are subject to theft and unauthorized copying for replay at a later
date.
A mobile computer visiting a new environment consumes some resources
of that environment. These resources includes network bandwidth, CPU
time, and disk storage of the base station. An accounting of this must be
made for both billing purposes and for limiting the impact of visiting
mobiles on the performance of a foreign environment.
Another security concern is how to prevent the visiting mobile computer
from abusing the foreign environment, in term of the bandwidth and the
disk storage it uses.

Asymmetry in communication
In a mobile computing environment, there is a difference in uplink and
downlink bandwidths due to the fact that mobile clients are not able to
transmit at very high speed, due to power restrictions and rather small
data transmission requirements of the mobile handsets. The bandwidth in
the downstream direction (server to mobile handsets) is far greater than
that in the upstream direction (handset to server).

6.3  Data Dissemination

Data dissemination deals with the delivery of data from a data repository
(server) to a large number of clients. The data dissemination model in a
mobile environment is usually different than its wired counterpart due to
the asymmetry in communication.
The model of sending data to clients in a mobile computing environment
is usually push-based. In a push-based data dissemination, data is sent to
the clients by the server without waiting for specific requests from the
client. In a push-based transmission, not only does the server save itself
from handling large numbers of similar queries from a large number
of clients, but also can propagate information whose existence may be
unknown to the clients.
On the negative side, push-based data dissemination unless judiciously
used can overwhelm the clients with unwanted information. Therefore, the
usefulness of a push-based data dissemination technique to a large extent
depends on the accuracy of predicting the requirements of various clients.
A simple solution usually used is to allow the clients to subscribe to specific
services or provide their profile, based on which the information can be
disseminated to them.
108  Fundamentals of Mobile Computing

6.4  Transaction Processing in Mobile Environment

In general, mobile transaction processing should provide seamless
transaction processing during disconnections, and preserve the system
correctness by allowing limited inconsistencies among the data copies.
It is important to recognize that clients can read or write data from any
one of the servers and move over to a new location and carry out further
operations from there. For example, a client can write at one server and
later read the item at a different server. The client can obtain inconsistent
views unless the different servers have been synchronized. The following
sections describe how each of the ACID properties has been redefined. The
relaxation of one property may impact the other ACID properties.

6.4.1  Atomicity Relaxation

In this modified scheme, an MH is allowed to submit ‘pieces’ of transactions
from different cells according to their movements. This approach weakens
the atomicity property and requires the ability to break a transaction into
sub-transactions that can be concurrently executed and interleaved with
the sub-transactions of other transactions while guaranteeing the ACID
properties.
Several techniques are available for decomposing a transaction into
sub-transactions according to different and at different levels of granularity
required and depending on the type of the transaction. That is, the read
only transactions may be submitted as a whole at a unique MSS or may be
split during processing as per the movements of the MH that submitted the
transaction. Transactions that update data items can be split into mutually
independent sub-transactions. This decomposition ensures that the split
sub-transactions can be concurrently executed at different MSSs and their
execution order does not impact the successful commitment of the original
transaction. The splitting in this case is done according to Bernstein’s
conditions. Each transaction, Si, has a write set Wi and a read set Ri. Every
two sub-transactions, Si and Sj, of a transaction T must satisfy the following
sets of conditions to guarantee their independence.
Wi  Wj = 
Ri  Wj = 
Wi  Rj = 
A runtime support must be designed to compute the decomposition
of transactions.

6.4.2  Consistency Relaxation

Under this approach, the database is logically partitioned into “clusters”,
based on either some semantic properties or location proximity. Data in
 Mobile Databases   109

the same cluster must be strictly consistent, whereas a “bounded” degree

of inconsistency is tolerated among the clusters, according to a relaxed
definition of consistency. In this approach, a mobile host can download
one cluster before disconnection and execute transactions on this data by
preserving ACID properties and after reconnection it can be copied to the
original database.

6.4.3  Isolation Relaxation

Some transaction models have been devised for mobile environments in

which the isolation property is not guaranteed. That is, the intermediate
results of a transaction can be observed by other concurrent transactions.
This is usually a side effect of the relaxation of other ACID properties.

6.4.4  Durability Relaxation

Durability of committed transactions is mainly affected by the possibility

of MHs autonomously operating on data. A disconnected MH can only
commit a transaction locally if this transaction does not conflict with other
transactions executed on the same host while the host is disconnected. On
reconnection, the transactions are globally executed on different MHs.

6.5  Data Replication

Data replication is the process of maintaining a defined set of data in more

than one location. In a mobile computing environment, data replication
is essential since this would allow the user to continue operating during
temporary disconnections and would also improve the transaction success
rate and transaction throughput. In order to achieve this, data must be
replicated at multiple places including partial replication of the relevant
data on the mobile database of the MH. An MH needs to have its own
complete or partial copy of the relevant data from the main database so
that it can process data locally. In general, data replication provides the
following two advantages.

Data availability:  Data is made available locally to the remote databases.

This avoids the need to connect to a single central database through slow
and less reliable connections.

Faster response times:  Replication improves the response times for data
requests since requests are processed on a local server. Further, local
processing decreases the workload at the central database server.
110  Fundamentals of Mobile Computing

The number of replications of a database in a mobile computing

environment can be large. There is not much issue in replicating the
read-only (or cached) copies, however, the key issue is in managing the
replication of the updateable (or core) copies. Of course, the number of core
copies need to be kept small to contain the overhead of synchronization.

6.6  Mobile Transaction Models

Mobile transactions access remote data through low bandwidth wireless

connections, and access local data in the disconnected mode. Reliability and
consistency management are the main goals of mobile transaction models.
Some existing models of mobile transaction management include Reporting
and Co-Transactions model (Chrysanthis, 1993), Kangaroo Transaction
model (Margaret Dunham et al., 1997), Clustering model (Pitoura and
Bhargava, 1994), Isolation-Only Transactions (Satyanarayan et al., 1990)
and Two-tier Transaction model (Gary Shih, 2002).

6.7  Rollback Process

Consider a transaction which has started operation, and the modification

of a database is in progress. Now if an error occurs due to mobility, then
all the transactions may fail. In this case, rather than leaving the database
with partial results, the transaction will have to undo the updates and leave
the database with older values that were available before the transaction
started. Such a reversal to the initial state is known as rollback process.

6.8  Two-phase Commit Protocol

In a mobile environment, during transaction processing, databases, and

computer networking, an appropriate commit protocol needs to be used
at the transaction commit stage. The two-phase commit protocol (2PC) is
a type of an atomic commitment protocol. It is a distributed algorithm
that coordinates all of a distributed atomic transaction to decide whether
to commit or abort the transaction. The protocol achieves its goal even
in cases of temporary system failure (involving either process, network
node, or communication failures), and is widely used (Bernstein et al.,
1987). The two phases of the algorithm are the commit–request phase, in
which a coordinator process takes the necessary steps for either committing
or aborting the transaction. In the commit phase, the coordinator decides
whether to abort the transaction.
 Mobile Databases   111

Note:  The two-phase commit (2PC) protocol should not be confused

with the two-phase locking (2PL) protocol, which is a concurrency control
protocol.

6.9  Query Processing

An efficient database system management should minimize the number of

uplink requests (queries). For this, the client needs to cache some commonly
used data in its local memory. Also in a service invocation by a mobile
handset (client), the location-specific results become important.

6.9.1  Location-dependent Querying

Several attempts have been made to integrate Global Positioning System

(GPS) information into queries to create location-dependent services.
Examples of such services include local news, train or flight information,
etc. In web browsing, the client browser may insert the location information
in the query to seamlessly obtain location-specific services.

6.9.2  Query Optimization

During a transaction with a database, queries typically require gathering

and updating information in a database. Efficient query processing is
achieved by using query decomposition and query optimization. A query
can be decomposed into a set of algebraic expressions by using the
following techniques:
• Appropriate analysis
• Normalization (conjunctives as well as disjunctives)
• Semantic analysis
During query optimization, a query is parsed and normalized. The
parser ensures that the language, say SQL, syntax is correct. Normalization
ensures that all the objects referenced in the query exist. Permissions are
checked to ensure that the user has the permission to access all tables and
columns in the query.
A query parser parses and translates a given query into an immediate
form such as a relational algebraic expression. The parser achieves this
based on both the syntax of the query and the semantics of the query. A
parse-tree of the query is constructed and then translated into a relational
algebra expression.
112  Fundamentals of Mobile Computing

BOX 6.3  Query decomposition operations

We explain the basic query decomposition operations using a few simple

examples:
Normalization:  Transform the query to a normalized form.
Example:  Consider the following query: Find the names of employees who
have been working on project P1 for 12 or 24 months?
• The query in SQL:
SELECT ENAME
FROM EMP, ASG
WHERE EMP.ENO = ASG.ENO AND
ASG.PNO = ‘‘P1’’ AND
DUR = 12 OR DUR = 24
• The qualification in conjunctive normal form:
EMP.ENO = ASG.ENO ^ ASG.PNO = ”P1” ^ (DUR = 12 v DUR = 24)
• The qualification in disjunctive normal form:
(EMP.ENO = ASG.ENO ^ ASG.PNO = ”P1” ^ DUR = 12) v
(EMP.ENO = ASG.ENO ^ ASG.PNO = ”P1” ^ DUR = 24)

Analysis:  Detect and reject ”incorrect” queries; possible only for a subset of
relational calculus. Example of the type of incorrect query: Checks whether the
attributes and the relation names of a query are defined in the global Schema
or Checks whether the operations on attributes do not conflict with the types
of the attributes, e.g., a comparison > operation with an attribute of type string.

Elimination of redundancy:  Simplify the query by eliminating redundancies,

e.g., redundant predicates. Redundancies are often due to semantic integrity
constraints expressed in the query language.

Semantical analysis:  Identify the incorrect queries. Example of semantically

incorrect query: Checks whether the components contribute in any way to the
generation of the result.

The purpose of the data analysis is to improve the speed of answering

queries on that data. Analysis produces summary information that can
either answer a query without consulting the data itself, or else modify the
query to a form that the data server will be able to process more quickly.
This query-modification operation using the knowledge of the data is
known as Semantic Query Optimization.
A mobile computing environment poses several challenges in query
processing. For example, as per the mobile-computing model, the route between
a pair of hosts may change over time. This may change the location-specific
results. So the mobile-computing model may directly affect the database query
processing. In the case of distributed query processing, the communication
costs play an important role in the query optimization process. The issue of
host mobility, therefore, complicates the optimization process.
 Mobile Databases   113

6.10  Recovery

A database recovery step gets the database back to the last stable state that
was reached before a failure. A transaction processing system (TPS) may fail
for many reasons. These reasons could be varied and may include, among
other things, battery wear out, system failure, human errors, hardware
failure, incorrect or invalid data, computer viruses, software application
errors, natural or man-made disasters. As it is not possible to prevent all
TPS failures, a TPS must be able to detect failures immediately after they
occur. For this, the TPS must be able to detect and correct errors when
they occur and restore the database to a stable state. A TPS recovery takes
place by a recovery manager, based on the backup, journal, and checkpoint
information. We explain these concepts in the following:

Journal:  A journal maintains an audit trail of transactions and database

changes. The transaction logs and the database change logs are maintained.
A transaction log records all the essential data for each transaction,
including data values, the time of transaction and the terminal number. A
database change log contains the before and after copies of records that
have been modified by transactions.

Checkpoint:  A checkpoint is a record of the state of the database system

at a specific time. A checkpoint record contains the necessary information
to restart the system. In case of a failure, it is possible to resume processing
from the most-recent checkpoint. Of course, the processing performed after
the last checkpoint would be lost. Therefore, checkpoints should be taken
frequently.

Recovery Manager:  A recovery manager is a program which restores the

database to a correct state from which the transaction processing activities
can be resumed, in case there is a failure.

SUMMARY

A mobile database is expected to handle the special issues that arise in

case of transactions invoked by mobile clients. The client communicates
with the server using a wireless connection. A mobile database is an
adaptation of a traditional database to accommodate many of the issues
that a mobile environment imposes, such as disconnections, low battery
power, etc. A cache is maintained to hold the frequently used data and
transaction results so that they are not lost due to connection failures.
We discussed a transaction model suitable for a mobile computing
application.
114  Fundamentals of Mobile Computing

FURTHER READINGS

Barbara, D., “Mobile Computing and Databases—A Survey,” IEEE

Transactions on Knowledge and Data Engineering, Vol. 11, No. 1,
pp. 108–117, January 1999.
Bernstein A., Philip, Vassos Hadzilacos, Nathan Goodman, Concurrency
Control and Recovery in Database Systems, Chapter 7, Addison Wesley
Publishing Company, ISBN 0-201-10715-5, 1987.
Chrysanthis, P.K., “Transaction Processing in Mobile Computing
Environments,” Proceedings of IEEE Workshop on Advances in Parallel
and Distributed Systems, pp. 77–82, 1993.
Dirckze, R.A., L. Gruenwald, “A Toggle Transaction Management Technique
for Mobile Multidatabases”, Proceedings of the CIKM 98, Bethesda, MD,
USA, pp. 371–377, 1998.
Gary Shih and Simon S.Y. Shhim, A Service Management Framework for
M-Commerce Applications, Kluwer Academic Publishers, the Netherlands,
2002.
Lu, E. and Y. Cheng, “Design and Implementation of a Mobile Database
for Java Phones,” Computer Standards and Interfaces, Elsevier, Vol. 26,
pp. 401–410, 2004.
Margaret H. Dunham, Abdelsalam Helal, and Santosh Balakrishnan, A
Mobile Transaction Model that Captures Both the Data and Movement
Behavior, Baltzer Science Publishers, 1997.
Pitoura, E. and B. Bhargava, “Building Information Systems for Mobile
Environments”. In Third International Conformation and Knowledge
Management, CIKM’ 94, Gatheburg, MD, USA, ACM, pp. 371–378,
November 1994.
Satyanarayan, M., J.J. Kistler, P. Kumar, M.E. Okasaki, E.H. Siegel, and
D.C. Steere, “Coda: A Highly Available File System for a Distributed
Workstation Environment”, IEEE Transaction on Computers, Vol. 39,
No. 4, pp. 447–459, April 1990.

EXERCISES
1. Explain the different issues that arise in a transaction processing
system having mobile clients for satisfactory operation of an
application. How are these issues resolved?
2. What do you mean by serializability of transaction executions? How
does a DBMS ensure that the transactions are serializable?
 Mobile Databases   115

3. What do you understand by concurrent transactions? Using one

example, explain why arbitrary concurrency in transaction processing
should not be allowed.
4. Briefly explain why wireless communication requires higher energy
compared to that in wired networks.
5. What is data hoarding in a mobile database environment? Explain
why is it required?
6. What is dozing in a mobile computing environment? Why does
dozing occur? What are its implications in mobile transaction
processing?
7. What do you mean by atomicity relaxation?
8. What is a mobile database? How is it different from a traditional
database?
9. What do you understand by data dissemination in a mobile
computing environment? Discuss how data dissemination can be
achieved in the mobile computing environment.
10. Explain the rollback process during the execution of a mobile
transaction. How is it different from a traditional database rollback?
11. Explain the two-phase commit process in the context of mobile
transactions.
12. How does a database handle failure that occurs during a mobile
transaction processing?
13. Explain why wireless transmission of data consumes more power
compared to transmission of data over a wired medium.
14. What problems might occur if a traditional database system is used
in a transaction processing system having mobile clients? Explain
your answer using suitable examples.
15. Why is data replication important in a mobile computing application?
Discuss the issues of data replication in mobile applications.
 Mobile Ad Hoc Networks
7
I n recent years, the mobile ad hoc networks are becoming very popular.
We discussed in Chapter 1 that a wireless LAN can be set up with the
help of certain fixed infrastructures such as access points and routers.
However, in many situations such as that in a remote village or on a
railway station no networking infrastructures may exist, yet the user of a
mobile device might still like to communicate with another user who might
be in the same area or elsewhere. In this situation, if the mobile devices
present in the area can somehow discover each other and collaborate to
set up a wireless network among themselves, then the communication can
become possible. However, it is necessary that such a network should be
self-configuring (set up and configured automatically without requiring
any user intervention), since the network configuration keeps changing due
to the movements of mobile devices. This is important since mobile ad hoc
networks are realized through short-lived links among the mobile devices.
Such temporary or ad hoc networks that are established and maintained on
the fly and work without the support of any form of fixed infrastructure
such as a base station, are known as Mobile Ad hoc NETworks (MANETs).
MANETs are traditionally defined as self-configuring networks set
up among the hand-held devices of mobile users. However, of late
several specialized MANETs such as Wireless Sensor Networks (WSNs)
and Vehicular Ad hoc Networks (VANETs) have emerged. Each of these
specialized ad hoc networks is suitable for a specific kind of application.
Being ad hoc networks after all, all these networks share some basic
characteristics. However, there exist significant differences among them
with respect to their operation, design, and applications. The focus of this
chapter is to discuss some essential aspects of MANET applications and
operations. Considering their widespread acceptance and use, we discuss
WSNs in the next chapter as a special type of MANET. The term mobile ad
hoc network (MANET) was proposed by IETF in the year 2002. In this text,

116
 Mobile Ad Hoc Networks   117

we shall use the terms mobile ad hoc networks (MANETs) and wireless
ad hoc networks interchangeably.
This chapter is organized as follows: In Section 7.1, we discuss the
basics of mobile ad hoc networks. In Sections 7.2 and 7.3, we present
the characteristics and applications of MANET. In Sections 7.4 to 7.7, we
discuss some basic issues pertaining to ad hoc networks. In Section 7.8,
we discuss various routing issues in MANETs. In Sections 7.9 and 7.10,
we discuss VANETs (Vehicular Ad hoc NETworks) as a special type of
MANET. In Sections 7.11 and 7.12, respectively, we discuss the various
security issues and attacks in MANETs.

7.1  A Few Basics Concepts

In the preceding chapters, we discussed various aspects of wireless

networks and the vital role played by networking infrastructures such
as hubs, routers, and base stations in their operation. However, such
networking infrastructures may not be available in many situations
such as a disaster-hit locality or a remote location as we have already
pointed out.

7.1.1  How Is an Ad Hoc Network Set Up without

the Infrastructure Support?

Let us try to imagine how a set of mobile devices can communicate with
each other in the absence of any form of fixed networking infrastructures
such as hubs, routers, base stations, etc. It is not hard to visualize that in
this situation, a network can be established through cooperation among
the devices themselves. In a simplistic realization of this concept, a mobile
device wanting to communicate can forward its packets to its neighbours,
and the neighbour nodes in turn can forward those to their neighbours, and
so on until the destination is reached. This essentially forms a simple ad
hoc network. Based on this idea, Fig. 7.1 illustrates a schematic model of a
simple ad hoc network comprising three mobile devices, named S, R, and
D. In this figure, suppose the mobile device S wants to communicate with
the device D. Assume that S and D are not within the transmission range
of each other and cannot directly communicate with each other. However,
they can take the help of node R to relay packets from each other. Observe
that Fig. 7.1 depicts a protocol stack to highlight that even though R is
primarily an independent device and not a networking infrastructure, yet
R is acting as some sort of a router operating at the network (or Internet)
layer to facilitate communication.
118  Fundamentals of Mobile Computing

Figure 7.1  A schematic model of a mobile ad hoc network.

7.1.2  Why Is Routing in a MANET a Complex Task?

In a wired network, a router determines the path that needs to be followed

by a packet based on the information contained within the IP address of
the destination, and uses this information to forward a packet towards
its destination. In an ad hoc network, such a simple and efficient routing
protocol is difficult to deploy. First, it is very difficult to have a global
identifier assigned to every node that would also indicate the route to the
node. This is because the nodes keep on moving and the identity would
also have to change, which would incur an inordinately large overhead.
Further, even if a route between a pair of nodes is somehow determined,
routes become quickly obsolete since they dynamically get built and also
get dissolved. Suppose a route has already been determined between a
source and a destination and packets are getting transmitted on this route.
Some nodes forming this route may move away even as the packets are
getting transmitted, thus disrupting the communication.
In a nutshell, we can say that in a MANET the topology of the network
and consequently the routes between different devices change dynamically
as nodes move away or fail. This is in contrast to any wired network. As
a result, packet routing is a critical and complex issue in MANETs and
a central topic in any study of MANETs. Consequently, a satisfactory
routing protocol for ad hoc networks should be able to cope with factors
such as link breakages, and ensure that messages get transmitted efficiently
and seamlessly even when nodes move or shut down due to hardware
malfunctioning, depletion of battery energy, etc.

7.2  Characteristics of Mobile Ad Hoc Networks

(MANETs)

There are several characteristics that distinguish a MANET from an

infrastructure-based network. Naturally, these characteristics must be
 Mobile Ad Hoc Networks   119

carefully considered while trying to improvise new network protocols or

extend the traditional network protocols for use in a MANET. A few of
these important characteristics of MANETs are described below.

Lack of fixed infrastructure:  Lack of any specific networking infrastructure

is possibly the most distinguishing characteristic of a MANET. In the
absence of any fixed networking infrastructure, a pair of nodes can either
communicate directly when they are in the transmission range of each
other, or they can communicate using a multi-hop communication that
gets set up through several devices located between them. Based on
this characteristic alone, cellular networks and wireless LANs cannot be
considered to be MANETs.

Dynamic topologies:  Since the devices in a MANET are allowed to move

arbitrarily, the network topology can change unpredictably. The rate of
topology change depends on the speed of movement of the mobile devices.
The speed of movement of a mobile device can vary greatly with the time
of the day and the specific MANET application being considered.

Bandwidth constrained, variable capacity links:  Wireless links have

significantly lower capacity than their wired counterparts. Further, factors
such as fading, noise, and interference can change the available bandwidth
of a wireless link arbitrarily with time. Consequently, the bandwidth of a
link can change arbitrarily with time.

Energy constrained operation:  The nodes in a MANET rely on battery

power. These batteries are small and can store very limited amounts of
energy. On the other hand, transmissions and processing required during
routing involve expenditure of substantial amount of energy causing the
batteries to get rapidly drained out, unless the routing protocol is carefully
designed. Therefore, energy conservation is usually considered to be an
important objective of MANET routing protocols.

Increased vulnerability:  MANETs are prone to many new types of

security threats that do not exist in the case of their wired counterparts.
Many of these threats arise due to the underlying wireless transmissions
and the deployment of collaborative routing techniques. Further, there are
increased possibilities of eavesdropping, spoofing, denial-of-service attacks
in these networks. Further, it is very difficult to identify the attacker since
the devices keep moving and do not have a global identifier. Besides,
nodes are vulnerable to capture and compromise. We discuss these security
threats in greater detail in Section 7.10.

Other characteristics:  Other distinguishing characteristics of a MANET

include a distributed peer-to-peer mode of operation, multi-hop routing,
and relatively frequent changes to the concentration of nodes over any
specific area.
120  Fundamentals of Mobile Computing

On account of the above mentioned characteristics of MANETS, a

suitable protocol for MANETs needs to operate under several constraints.
We point out these constraints in the following Subsection 7.2.1.

7.2.1  MANET Operational Constraints

The nodes in a MANET have low processing capabilities and these are
connected by low bandwidth wireless links. To cope with low processing
capabilities of the nodes and the limited bandwidth availability, a suitable
routing protocol needs to be adopted to make efficient utilization of the
available bandwidth. However, an appropriate routing protocol for a
MANET should keep the computational and communicational overheads
low, since the nodes in a MANET have low computational capability,
storage capacity and battery power. Out of these, battery power is possibly
the most scarce resource. To conserve communication resources, a suitable
routing protocol for a MANET should not make use of frequent flooding
or even make use of periodic updated messages. Since nodes in a MANET
are likely to be mobile, in case a route in use dissolves due to the rapidly
changing topology, a routing protocol for an ad hoc network should be
able to find an alternate route very quickly.

7.3  Applications of MANETs

There is at present a widespread interest in MANETs among practitioners

and researchers due to the significant advantages they offer for certain
types of applications. A MANET can be set up quickly since no fixed
infrastructures need to be deployed. Thus, in any situation where fixed
infrastructure becomes difficult to be set up because of security, cost,
inaccessibility of the terrain, or safety-related reasons, ad hoc networks
become the preferred choice. Of the large number of applications that
are possible with MANETs, a few example applications are defence-
related operations and disaster management applications. Some example
applications for which ad hoc networks are being used are described below.

Communication among portable computers

Miniaturization has allowed the development of many types of portables
and computerized equipment, which have become very popular. Many
of these portables work meaningfully when connected to some network,
possibly a LAN or the Internet. For this, the portables are typically required
to be within the range of some wireless hub. Satisfaction of this requirement
would, however, drastically reduce the flexibility and the mobility of the
devices. As an example, consider a lecture room where no networking
infrastructures exist. In this case, using MANET the audience can exchange
 Mobile Ad Hoc Networks   121

notes, and also can surf the Web if at least one of the hand-held devices
has access to Internet, for example, through a data card. If the mobile
devices are present in sufficient density, network connections among them
can be established seamlessly to form a MANET over which the nodes can
communicate and carry out the network operations.

Environmental monitoring
A popular category of applications of MANETs is the collection of the
various types of data about the environment in which they are deployed.
Continuous data collection from remote locations is considered important
for several applications such as environmental management, security
monitoring, road traffic monitoring and management, etc. Miniaturized
sensors have proved to be an effective means of gathering environmental
information such as rainfall, humidity, presence of certain animals, etc.
As planting sensors in the environment is becoming common, ad hoc
networking is gaining more ground and becoming more prevalent. In this
environmental monitoring application, a large number of sensors nodes
are deployed in the environment. Such ad hoc sensor networks can be
deployed to collect data from remote locations and the sensor nodes can
even respond to some commands issued by the data collection centre.
MANETs efficiently handle the introduction of new sensors into an already
operational sensor network as well as can handle dynamic disconnections
of nodes. Since each sensor acts as a hub, the range over which the sensors
can be spread is tremendously increased, because they do not have to be
deployed around some central station. For this application, use of energy
efficient protocols becomes imperative since it can help increase the life
span of the network.

Military
The present-day military equipment have become quite sophisticated,
have many automated parts and contain one or more computers. This
opens up the scope of setting up an ad hoc network consisting of various
military equipment deployed in a frontline battle field. Ad hoc networking
of these equipment can allow a military setup to take advantage of an
information network among the soldiers, vehicles, and military information
headquarters. For example, an ad hoc network can be automatically set up
at a battlefront among the equipment, and the hand-held devices can collect
information from and disseminate command to the frontline personnel.

Emergency applications
Ad hoc networks do not require any pre-existing infrastructure. These
networks, therefore, can be deployed easily and rapidly in emergency
situations such as a search and rescue operation after a natural disaster,
and for applications such as policing and fire fighting. In these situations,
122  Fundamentals of Mobile Computing

ad hoc networks can be set up on the fly. As an example, consider the

situation where a severe earthquake has hit a locality. All the commu-
nication infrastructure would be destroyed and it would not be possible
to immediately re-establish communication using the traditional
infrastructure-based networks such as telephones, mobile phones, Internet,
etc. In such a scenario, an ad hoc network can be quickly set up to provide
network connectivity to rescue personnel in order to facilitate the rescue
operation.

7.4  MANET Design Issues

Before we discuss some important MANET protocols, we point out below

a few important issues that are relevant to the design of suitable MANET
protocols.

Network size and node density

Network size and node density are the two important parameters of
a MANET that need to be considered while designing an appropriate
routing protocol for a network. Network size refers to the geographical
coverage area of the network and network density refers to the number of
nodes present per unit geographical area. For larger networks, clustering
is essential to keep the communication overheads low. The cluster size as
well as a specific clustering solution for a network would, to a large extent,
depend on node density.

Connectivity
The term connectivity of a node usually refers to the number of neighbours
it has. Here a neighbour of a node is one that is in its transmission range.
The term connectivity is also sometimes used to refer to a link between
the two nodes. The term link capacity denotes the bandwidth of the link.
In a MANET, both the number of neighbouring nodes and the capacities
of the links to different neighbours may vary significantly.

Network topology
The topology of a network denotes the connectivity among the various
nodes of the network. Mobility of the nodes affects the network topology.
Due to node mobility, new links can form and some links may get
dissolved. Other than mobility, nodes can become inoperative due to
discharged batteries or hardware failures, and thereby cause changes to the
topology. The rate at which the topology changes needs to be appropriately
considered in the design of an effective network.
 Mobile Ad Hoc Networks   123

User traffic
The design of a MANET is carried out primarily based on the anticipated
node density, average rate of node movements, and the expected traffic.
The traffic in a network can be of various types. A network protocol
should leverage the characteristics of specific traffic types that are expected
to improve its performance. The common traffic types are the following:
• Bursty traffic
• Large packets sent periodically
• Combination of the above two types of traffic

Operational environment
The operational environment of a mobile network is usually either urban,
rural and maritime. These operational environments support the Line of
Sight (LOS) communication. But, there can be a significant difference in
the node density and mobility values in different operational environments,
requiring different designs of mobile networks to suit an operational
environment.

Energy constraint
As already mentioned, no fixed infrastructure exists in a MANET; the
mobile nodes themselves store and forward packets. This additional role
of mobile nodes as routers leads to nodes incurring perennial routing-
related workload and this consequently results in continual battery
drainage. Though this overhead is indispensable if the network is to be
kept operational, the energy spent can be substantially reduced by allowing
the nodes to go into a sleep mode whenever possible.

7.5  Routing

As we have already discussed, packet routing is usually a much more

complex task in an ad hoc network compared to that of an infrastructure-
based network. The main complications arise on account of continual
topology changes and limited battery power of the nodes. Recall that
we discussed these issues in Section 7.4 and a few other issues that are
inherent to MANETs. When the destination node is not in the transmission
range of the source node, the route has to be formed with the help of the
intervening nodes in the network.
As we know, the purpose of routing is to find the best path between
the source and the destination for forwarding packets in any store-and-
forward network. In a traditional network, routing is a relatively easy task
because the routes to nodes can be uniquely and efficiently identified based
124  Fundamentals of Mobile Computing

on the subnet structure encoded in IP. In a MANET, the nodes making up

a route may themselves move or shut down due to low battery energy, in
the process making the knowledge about routes at various nodes to quickly
become obsolete. It is therefore necessary to find a new route each time
a node needs to transmit a message, making routing an expensive and
difficult task.
Based on the above discussions, it may be seen that:
• Traditional routing protocols would not be suitable in an ad hoc
network.
• Each node in an ad hoc network needs to have routing capability and
also needs to participate in routing to keep the network operational.
We can now state that whenever there is an incoming packet in a
MANET:
(a) Forward the packet to the next node (hop).
(b) While forwarding the packet, the sender needs to ensure that:
(i) The packet moves towards its destination.
(ii) The number of hops/path length is minimized.
(iii) Delay is minimized.
(iv) The packet loss is minimized.
(v) The packet does not move around the network endlessly.
Several types of routing protocols have been proposed for MANETs.
Different routing protocols essentially implement the above steps (a) and
(b) while meeting the constraints inherent to the network, such as low
energy consumption, through the deployment of various techniques. We
will now review the essential concepts of a traditional routing technique.
Later, we will build upon these concepts to introduce the routing protocols
for ad hoc networks. No simple IP-address based routing is possible in
a MANET due to the continual topology changes on account of node
movements.

7.6  Essentials of Traditional Routing Protocols

Before understanding the MANET routing protocols, it is necessary to

have a clear understanding of the routing mechanisms deployed in a
traditional network. It will help us appreciate the specific changes made
to traditional routing protocols to support the specific requirements of an
ad hoc network. Two important classes of routing protocols for traditional
networks are the link state and the distance vector. These two protocols
are extremely popular in packet-switched networks. Both these protocols
require a node to determine the next hop along the “shortest path” towards
a given destination. The shortest path is computed according to some
specific cost metric such as the number of hops in the route.
 Mobile Ad Hoc Networks   125

7.6.1  Link State Protocols (LSP)

The term link state denotes the state of a connection of one router with one
of its neighbours. A neighbour of a router is one with which it can directly
communicate without taking any help from the intervening routers. Each
router determines its local connectivity information, and floods the network
with this information with a link state advertisement. As a router in the
network receives this link state advertisement, it stores this packet in a link
state packet database (LSPDB). This storage of link state advertisements in
an LSPDB is in addition to the routing table that each router maintains. It
is easy to see that all routers in the network will have identical LSPBDs.
Based on the bits and pieces of information stored in its LSPDB, each router
constructs the connectivity information for the entire network as a graph
using the Dijkstra’s shortest path algorithm. Once a router constructs this
graph, it computes the routing table from this and uses it in all its routing
decisions. Thus, we can say that a router in the link state protocol bases
its routing decisions on messages (link state advertisements) received from
other routers in the network regarding their link states or the state of its
connectivity with other routers.
A basic characteristic of a link state routing protocol is that every router
constructs a graph representing the connectivity between the various nodes
in the network based on the information received from other routers. We
can think of this construction of the map of the entire network from bits
and pieces of information received from other routers as similar to the
solution of a zig-saw puzzle by putting together the different pieces of
the puzzle. This graph representing the network is usually constructed in
the form of a tree with the local router forming the root of the tree. The
graph captures the shortest path route from the root to any other router.
Once a node constructs this tree, it computes the best path from itself to
every other node in the network and stores this information in the form of
a routing table. This contrasts with the distance-vector routing protocols,
in which each router shares its routing table with its neighbours. But in a
link state protocol, only the connectivity related information is exchanged
between routers, and no complete routes are exchanged.
In a link state protocol, each router periodically determines the state
of its links to its neighbours by exchanging hello packets with them across
all its network interfaces. Figure 7.2 shows how a router is connected to
other routers through links established by its network interfaces. Based on
the reply received from its neighbours, the router determines the state of
the link in terms of the delay and other characteristics. Subsequently, the
router forms a short message called the link state advertisement and sends it
to its neighbours. A link state advertisement is also sent whenever a router
experiences a connectivity change. A link state advertisement message
contains:
• The identity of the router originating the message.
• The identities of all its neighbours.
126  Fundamentals of Mobile Computing

• The delays along various links to its neighbours.

• A unique sequence number, which is formed by increasing the count
every time the router forms a new link state advertisement.

Figure 7.2  Schematic diagram of a router.

This link state advertisement is then flooded throughout the network

as follows: A router sends a copy of a link state advertisement to all of
its neighbours. A router receiving this message examines the sequence
number of the last link state advertisement from the originating router
by consulting its LSPDB. If this received link state advertisement is more
recent, it replaces the last message with the currently received message in
its LSPDB, and also forwards a copy of this link state advertisement to
each of its neighbours.
Using the latest link state advertisements stored in its LSPDB, a router
can easily reconstruct the network topology in the form of a tree by using
the Dijkstra’s iterative shortest path algorithm that we discuss below. This
algorithm constructs the shortest path tree edge by edge, at each step
adding one new edge, corresponding to the construction of the shortest
path to a router. During the construction of the link state tree, if there are
any inconsistencies among the reported link state advertisements from
different routers, then the same have to be resolved. If one router reports
that it is connected to another, but the other node does not report that it
is connected to the first, then that link is not included on the tree.
Let us now discuss the construction of the link state tree for a router
from the link state advertisements stored in its LSPDB. A router maintains
two data structures: a tree containing nodes which are done, and a list
of candidates. This tree is essentially a shortest path first (SPF) tree. The
algorithm starts with both data structures empty. It first adds itself to the
tree and thus is at the root. Then a greedy iterative algorithm based on the
Dijkstra’s algorithm is deployed to repetitively carry out the following by
examining the link state advertisements that a router received in the past.
All routers which are connected to the router just added to the tree,
excepting any routers which are either already in the tree or in the candidate
list, are added to the candidate list.
 Mobile Ad Hoc Networks   127

The delays from each router in the candidate list to every other router
in the tree are compared. The candidate router having the shortest delay
is moved into the tree and attached to the appropriate neighbour router.
Whenever a router is moved from the candidate list into the tree, it is
removed from the candidate list.
The above two steps are repeated till there are no more routers left
in the candidate list. If different routers somehow have maps that are
inconsistent, then routing loops can form. In the simplest case of formation
of a routing loop, two neighbouring routers determine that the other is the
best path to a given destination, and therefore packets traverse endlessly
between them. Routing loops involving more than two nodes are also
possible.
The reason for the formation of routing loops is fairly simple. Each node
computes its shortest-path tree and its routing table without interacting in
any way with any other node. Therefore, if two nodes start with different
maps, it is easy to have scenarios in which routing loops involving multiple
nodes are created. Examples of these situations can be easily constructed
and hence are not discussed here.
Once the network topology has been determined in the form of a
shortest path tree, a router forms its routing table and uses it to find the
best route to any destination. Based on this, it can determine the optimal
next hop for each destination in the network. Two widely used link state
protocols in traditional networking are OSPF (Open Shortest Path First)
and IS-IS (Intermediate System To Intermediate System).

7.6.2  Distance Vector (DV) Protocols

The distance vector protocols get their name from the fact that they base
their routing decisions on the distance to the destination in terms of the
number of hops that a packet will have to traverse to reach its destination.
You might wonder what exactly the term vector denotes. The term vector
here means that routes are advertised as a vector (distance, direction),
where distance is the number of hops between the two nodes and direction
is defined in terms of the next hop router to which the packets need to
be forwarded. The distance vector protocols are based on the well-known
Bellman-Ford algorithm.
The distance vector routing protocols share everything they know about
the various routes in the network with their neighbours by broadcasting
their entire route table. Each node advertises its entire routing table to its
immediate neighbours only. A router transmits its routing table that has
been formed from its own perspective. That is, it represents the routes
to various routers from itself. For example, “router A is at a distance
of five hops away, in the direction of the neighbour router X.” As this
statement implies, each router learns routes from its neighbouring routers’
perspectives, and based on this forms its own perspective of the routes and
128  Fundamentals of Mobile Computing

then advertises the routes from its own perspective. As a router receives the
routing information of its neighbouring nodes, it updates its own routing
table by examining the received information and in turn informs its own
neighbours of the changes. This is also referred to as “routing by rumour”
because routers are relying on the information they receive from other
routers and have no way to determine if the information is actually valid.
A number of techniques have been suggested to cope with instability and
inaccurate routing information.
As we have seen, the routers using the distance vector protocol do not
have knowledge of the entire path that a packet would take to reach its
destination. Instead, they just know the following vector:
1. Direction in (or the specific network interface over) which a packet
should be forwarded.
2. Its own distance from the destination.
The DV protocol is, therefore, based on calculating the distance and
the direction to any router in a network. The cost of reaching a destination
is calculated using the various route metrics. The two popular distance
vector routing protocols are RIP (Routing Information Protocol) and
IGRP (Interior Gateway Routing Protocol). RIP uses the hop count of the
destination whereas IGRP takes into account the other information such
as node delay and available bandwidth. RIP supports the cross-platform
distance vector routing, whereas IGRP is a Cisco Systems proprietary
distance vector routing protocol.
As we have already mentioned, the distance vector routing protocol
requires each router to periodically send its entire route tables to all its
neighbours. A notable exception to this convention is the Cisco’s Enhanced
IGRP (EIGRP). Though EIGRP is a distance vector protocol, it does not
require transmitting updates periodically. Further, the updates are not
broadcast and do not contain the full route table.

7.7  Routing in MANETs: A Few Basic Concepts

In this section, we present a few basic concepts and definitions, based on

which we will discuss the MANET routing protocols. We first highlight a
few important differences of MANET routing from the traditional routing
protocols that are deployed in wired networks. This will help us to better
appreciate the routing protocols in MANETs.

7.7.1  Routing in MANETs vs. Routing in Traditional Networks

The following are the three important ways in which a MANET routing
protocol differs from routing of packets in a traditional network.
 Mobile Ad Hoc Networks   129

• In a MANET, each node acts as a router, whereas ordinary nodes in

a traditional wired network do not participate in routing the packets.
• In a MANET, the topology is dynamic because of the mobility of
the nodes, but it is static in the case of traditional networks. Thus,
the routing tables in a MANET quickly become obsolete, making
the routing process complicated.
• In the simple IP-based addressing scheme deployed in wired
networks, the IP address encapsulated in the subnet structure does
not work because of node mobility.
In order to cope with the above three important differences, each
MANET node needs to carry out the following important routing tasks:
route discovery and route maintenance. We explain these tasks in the
subsequent sections.

Types of communications
In a network, a node can initiate the following types of communications:
Unicast:  In this, a message is sent to a single destination node.
Multicast:  In this type of transmission, a message is sent to a selected
subset of the network nodes.
Broadcast:  In this type of transmission, a message is sent to all the
nodes in the network. Since unrestrained broadcast communications can
choke a MANET, applications usually do not use broadcast communication.
As we shall see, different protocols support different types of
communications with different levels of proficiency. The exact routing
protocol to be deployed in a network is determined based on the results
of a traffic analysis characterizing the transmissions in the network into
unicast, multicast, and broadcast. Since the broadcast transmission is
normally not deployed in a MANET, the routing protocols in a MANET
can broadly be classified into unicast and multicast types.

7.7.2  A Classification of Unicast MANET Routing Protocols

Unicast routing protocols in MANETs are classified into proactive (table-
driven), reactive (on-demand) and hybrid protocols. This classification is
based on how a protocol manages to determine the route correctly in the
presence of topology changes.

Proactive (table-driven) protocols

A proactive routing protocol is also known as a table-driven routing protocol.
In this protocol, each node in a routing table maintains information about
routes to every other node in the network. These tables are periodically
updated in the face of random network topology changes. An example of
130  Fundamentals of Mobile Computing

a proactive (table-driven) protocol is the Destination Sequenced Distance

Vector (DSDV) protocol. Since each node knows the complete topology, a
node can by itself determine the best route to a destination based on its local
information. Since the topology can change due to node movements and
nodes shutting down, the routing table needs to be updated periodically.
Therefore, a proactive protocol usually generates a large number of control
messages to keep the routing tables up-to-date at different nodes. This
routing overhead may take up a large part of the available bandwidth.
Especially for networks with a large number of nodes and very high node
mobility, the control messages may consume almost the entire bandwidth.
Obviously, this protocol is not suitable for large networks as the size of
the routing table may become excessively large due to a node maintaining
routes to each and every other node in the network.

Reactive (on-demand) protocols

A reactive routing protocol is also known as an on-demand routing
protocol, since in this protocol nodes do not maintain up-to-date routes to
different destinations, and new routes are discovered only when required.
When a node does not have knowledge about any route to a specific
destination, it uses a flooding technique to determine the route. These
protocols were designed to reduce the large overheads incurred by the
proactive protocols. This efficiency is achieved by maintaining information
for active routes only and doing away with a large number of route update
messages. Two examples of on-demand routing protocols are:
(i) Dynamic source routing (DSR)
(ii) Ad hoc on-demand distance vector routing (AODV)

Hybrid routing protocols

Hybrid routing protocols have the characteristics of both proactive and
reactive protocols. These protocols combine the good features of both the
protocols. The hybrid routing protocols are designed to achieve increased
scalability by allowing nodes with close proximity to work together to form
some sort of a backbone to reduce the route discovery overheads. This is
mostly achieved by proactively maintaining routes to nearby nodes and
determining routes to far away nodes only when required using a route
discovery strategy. Most hybrid protocols proposed to date are zone-based,
which means that the network is partitioned or seen as a number of routing
zones by each node. An example of a hybrid routing protocol is the Zone
Routing Protocol (ZRP).

7.8  Popular MANET Routing Protocols

Based on the classification of the routing protocols discussed in Section

7.7.2, we now discuss a few popular MANET routing protocols.
 Mobile Ad Hoc Networks   131

7.8.1  Destination-Sequenced Distance-Vector Routing Protocol

Destination-Sequenced Distance-Vector Routing (DSDV) is an important

MANET routing protocol. It is based on the table-driven (proactive)
approach to packet routing (Barbora, 1999). It extends the distance vector
protocol of wired networks just as the traditional algorithm makes uses of
the classical Bellman–Ford routing algorithm. An improvement made here
is the avoidance of routing loops through the use of a number sequencing
scheme. In DSDV, each node in a MANET maintains a routing table in
which all of the possible destinations and the number of hops to each
destination are recorded. Hence, routing information is always readily
available, regardless of whether the source node requires a specific route
or not.
Each node maintains information regarding routes to all the known
destinations. The routing information is updated periodically. This can be
considered a shortcoming of the protocol since it deprives a node from
going into sleeping mode. Also, there is a traffic overhead even if there
is no change in network topology. Further, nodes maintain routes which
they may never use.
A sequenced numbering system is used to allow mobile nodes to
distinguish stale routes from new ones. Updated routing tables are
exchanged periodically among the nodes of the network to maintain
table consistency. A naïve table exchange approach would generate a
lot of control traffic in the network, leading to inefficient utilization of
the network resources. To alleviate this problem, DSDV uses two types
of route update packets. The first is known as full dump. This type of
packet carries all the available routing information and can require
multiple network protocol data units (NPDUs) to be transmitted. During
periods of occasional movement, these packets are transmitted infrequently.
Smaller incremental packets are used to disseminate only the information
that has changed since the last full dump. These incremental broadcasts
usually fit into a standard NPDU, thereby decreasing the amount of traffic
generated. The mobile nodes maintain an additional table where they store
the data received through the incremental routing information packets
from various nodes.
New route broadcasts contain the address of the destination, the
number of hops to reach the destination, as well as a unique sequence
number. The route labelled with the most recent sequence number is
always used. Up on a change, a node might receive several messages from
different sources. The weighted average time that routes to a destination
will fluctuate before the route with the best metric received. Mobiles
also keep track of this settling time of routes. By delaying the broadcast
of a routing update by the length of the settling time, mobiles can
reduce network traffic and optimize routes by eliminating the suboptimal
ones.
132  Fundamentals of Mobile Computing

Important steps in the operation of DSDV

The important steps in the operation of DSDV are summarized below:
1. Each router (node) in the network collects route information from
all its neighbours.
2. After gathering information, the node determines the shortest path
to the destination based on the gathered information.
3. Based on the gathered information, a new routing table is generated.
4. The router broadcasts this table to its neighbours. On receipt by
neighbours, the neighbour nodes recompute their respective routing
tables.
5. This process continues till the routing information becomes stable.
DSDV incorporates a sequenced numbering scheme. Each routing
advertisement comes with a sequence number. Within an ad hoc network,
advertisements may propagate along many paths. Sequence numbers help
a node to consider the advertisements in the correct order. This avoids the
loops that may form while using the unchanged distance vector algorithm.
Figure 7.3 shows an example of a MANET. Table 7.1 is the routing table
of the node N4 at the moment before the movement of nodes . The metric
field in the routing table helps to determine the number of hops required
for a packet to traverse to its destination. The install time indicates when
the entry was made. It is used to delete stale entries from the table.

N1 N3
N2

N4 N5

Figure 7.3  An example of a MANET topology at a given instant of time.

TABLE 7.1  DSDV Routing Table for the MANET of Figure 7.3 at Node N4
Destination Next hop Metric Sequence no. Install time
N1 N1 1 321 001
N2 N2 1 218 001
N3 N2 2 043 002
N5 N5 1 163 002

7.8.2  Dynamic Source Routing (DSR) Protocol

Dynamic Source Routing (DSR) (Johnson and Maltz 1996, 2001) protocol
was developed to be suitable for use in a MANET having a reasonably
small diameter of about 5 to 10 hops and when the nodes do not move very
 Mobile Ad Hoc Networks   133

fast. DSR is a source initiated on-demand (or reactive) routing protocol for
ad hoc networks. It uses source routing, a technique in which the sender
of a packet determines the complete sequence of nodes through which a
packet has to travel. The sender of the packet then explicitly records this
list of all nodes in the packet’s header. This makes it easy for each node in
the path to identify the next node to which it should transmit the packet
for routing the packet to its destination.
In this protocol, the nodes do not need to exchange the routing table
information periodically, which helps to reduce the bandwidth overhead
associated with the protocol. Each mobile node participating in the protocol
maintains a routing cache which contains the list of all routes that the node
has learnt. Whenever a node finds a new route, it adds the new route to its
routing cache. Each mobile node also maintains a sequence counter called
request id to uniquely identify the last request it had generated. The pair
< source address, request id > uniquely identifies any request in the ad
hoc network. DSR works in two phases: (i) Route discovery and (ii) Route
maintenance. We discuss these two phases in the following.

Route discovery
Route discovery allows any host to dynamically discover the route to any
destination in the ad hoc network. When a node has a data packet to
send, it first checks its own routing cache. If it finds a valid route in its
own routing cache, it sends out the packet using this route. Otherwise, it
initiates a route discovery process by broadcasting a route request packet
to all its neighbours. The route request packet contains the source address,
the request id and a route record in which the sequence of hops traversed
by the request packet, before reaching the destination is recorded.
A node upon getting a route request packet does the following. If a
packet does not have the required route in its routing cache, it forwards
the packet to all its neighbours. A node forwards a route request message
only if it has not yet seen it earlier, and if it is not the destination. The
route request packet initiates a route reply upon reception either by the
destination node or by an intermediate node that knows a route to the
destination. Upon arrival of the route request message at the destination,
this information is piggybacked on to the route reply message that contains
the path information and is sent to the source node.
The route discovery process is schematically shown in Figure 7.4. As
shown in the figure, suppose a node N1 wishes to send a message to the
destination node N8. The intermediate nodes are N2, N3, N4, N5, N6, N7.
The node N1 initiates the route discovery process by broadcasting a route
request packet to its neighbours N2 and N3. Note that each node can have
multiple copies of the route request packet arriving at it. The propagation
of route reply is shown in Figure 7.5, and the acknowledgement messages
from destination to source are indicated by thick arrows.
134  Fundamentals of Mobile Computing

N1-N2 N1-N2-N5 Destination

N1 N2 N8
N5
Source
N1-N3-N4-N7
N1-N3-N4
N1
N1-N3-N4
N7
N1 N4
N1-N3 N1-N3-N4-N6
N1-N3-N4
N3 N6

Figure 7.4  An example of the route discovery process in DSR.

N1-N2-N5-N8
N1-N2-N5-N8 Destination
N1-N2-N5-N8 N2 N8
N5
Source

N1

N7
N4

N3 N6

Figure 7.5  An example of the propagation of route reply in DSR.

Route maintenance
A known route can get broken either due to the movement of some
nodes making up the route or the battery of a node forming part of the
route getting exhausted. Route maintenance is the process of monitoring
the correct operation of a route in use and taking any corrective action
when needed. When a host (source) while using a route, finds that it is
inoperative, it carries out route maintenance. Whenever a node wanting to
send a message finds that the route is broken, it would help if it already
knows of some alternative routes. Since the nodes do not exchange any
routing information in this protocol, whenever a node detects that one of
its next hop neighbour node is not responding, it sends back a route error
packet containing its own address and the address of the hop that is not
working. As soon as the source node receives the RouteError message, it
deletes the broken-link-route from its cache. If it has another route to the
destination, it starts to retransmit the packet using the alternative route.
Otherwise, it initiates the route discovery process again.
 Mobile Ad Hoc Networks   135

7.8.3  Ad Hoc On-demand Distance Vector (AODV)

The route discovery and route maintenance activities in AODV (Perkins

et al., 2001) are very similar to those for the DSR protocol. AODV does
make use of hop-by-hop routing, sequence numbers and beacons. The
node that needs a route to a specific destination generates a route request.
The route request is forwarded by intermediate nodes which also learn a
reverse route from the source to themselves. When the request reaches
a node with route to destination, it generates a route reply containing
the number of hops required to reach the destination. All nodes that
participate in forwarding this reply to the source node create a forward
route to destination. This route created from each node from source to
destination is a hop-by-hop route. Recollect that DSR includes the complete
route in packet headers. The large headers can substantially degrade the
performance, especially when the data content of packets is small. AODV
attempts to improve upon DSR by maintaining routing tables at the
nodes, so that the data packets do not have to contain the routes. AODV
retains a positive feature of DSR, in that the routes are maintained only
between those nodes that need to communicate. If a link break occurs
while a route is being used to transmit a message, a route error message
is sent to the source node by the node that observes that the next link in
the route has failed.

7.8.4  Zone Routing Protocol

The Zone Routing Protocol (ZRP) (Haas, 1997) is a hybrid protocol. It

incorporates the merits of both on-demand and proactive routing protocols.
A routing zone is similar to a cluster. However, unlike clusters, zones
can overlap. A routing zone comprises a few MANET nodes within a few
hops from the central zone. Within a zone, a table-driven routing protocol
is used. This implies that regular route updates take place only within
a zone. Each node, therefore, has a route to all other nodes within the
zone. If a destination node happens to be outside the source’s zone, ZRP
employs an on-demand route discovery procedure which works as follows.
The source node sends a route request to the border nodes of its zone,
containing its own address, the destination address and a unique sequence
number. Border nodes are those nodes which are some predefined number
of hops away from the source. Each border node checks its local zone for
the destination. If the destination is not a member of its local zone, then
the border node adds its own address to the route request packet and
forwards the packet to its own border nodes. When the destination node
is reached in this process, a route reply is sent on the reverse path, back to
the source. The source node uses the path saved in the route reply packet
to send data packets to the destination.
136  Fundamentals of Mobile Computing

7.8.5  Multicast Routing Protocols for MANET

As we have already discussed, multicast is the delivery of a message to

a group of destination nodes in a single transmission as shown in Figure
7.6. For efficient operation of a multicast routing protocol, it is necessary
to minimize the unnecessary packet transmissions as well as minimize
the energy consumption. In order to achieve this, a multicast transmission
should not be approximated by multiple unicast transmissions.

Figure 7.6  Multicast transmission.

Efficient multicast routing is much more difficult to achieve in a

MANET compared to any other network. This is due to host mobility,
broadcast nature of wireless environment, and interferences from various
noise sources. The proposed MANET multicast routing protocols either
modify the conventional tree structure, or deploy a different topology
between group members. The popular MANET multicasting protocols are
either tree-based or mesh-based:

Tree-based protocol
Tree-based schemes establish a single path between any two nodes in
the multicast group. These schemes require minimum number of copies
per packet to be sent along the branches of the tree. Hence, they are
bandwidth efficient. However, as mobility increases, link failures trigger
the reconfiguration of the entire tree. Another complicacy is that when
there are many sources, a node either has to maintain a shared tree, losing
path optimality, or maintain multiple trees resulting in storage and control
overhead. An examples of this category of protocol is the Multicast Ad Hoc
On-demand Distance Vector (MAODV) (Royer and Perkins, 1999) Routing
Protocol.

Mesh-based protocol
Mesh-based schemes establish a mesh of paths that connect the sources and
destinations. They are more resilient to link failures as well as to mobility.
The major disadvantage of this scheme is that multiple copies of the same
packet are disseminated through the mesh, resulting in reduced packet
 Mobile Ad Hoc Networks   137

delivery and increased control overhead under highly mobile conditions.

An examples of this category of protocol is the On-demand Multicast
Routing Protocol (ODMRP) (Lee et al., 2002).

7.9  Vehicular Ad Hoc Networks (VANETs)

A Vehicular Ad Hoc Network (VANET) is a special type of MANET in

which moving automobiles form the nodes of the network. VANETs were
initially introduced for vehicles of police, fire brigades, and ambulances
for safe travelling on road. In this network, a vehicle communicates with
other vehicles that are within a range of about 100 to 300 metres. Multi-hop
communication often results in rather large networks. In a city or a busy
highway, the diameter of the network can be several tens of kilometres.
Any vehicle that goes out of the signal range of all other vehicles in the
network is excluded from the network. A vehicle that was outside the
communication range of all other vehicles of a VANET can come in the
range of a vehicle that is already in the network and as a result can join
the network.

Figure 7.7  A VANET use scenario.

A VANET can offer a significant utility value to a motorist. A few

important uses of a VANET are described below:
• A VANET can help drivers to get advance information and warnings
from a nearby environment via message exchanges. For example,
in the scenario shown in Figure 7.6, two vehicles are involved in a
collision. The trailing vehicles get advance notification of the collision
ahead on the road. The driver can also get advance information on
the road condition ahead, or a warning about the application of
emergency electronic brake by a vehicle ahead in the lane.
• A VANET can help disseminate geographical information to the
driver as he continues to drive. For example, the driver would be
notified of the nearby food malls or petrol refilling stations, map
display, etc.
138  Fundamentals of Mobile Computing

• Drivers may have the opportunity to engage in other leisurely tasks,

such as VoIP with family, watch news highlights, listen to series
of media files known as podcasts, or even carry out some business
activities such as participate in an office video conference session.

7.10  MANET vs VANET

A MANET, as we have already defined, is a collection of mobile nodes

that communicate with each other over bandwidth constrained wireless
links without any infrastructure support. In this sense, we can consider
a VANET to be a special category of MANET. The nodes are mobile in
VANETs as well as in MANETs. However, the VANET nodes (vehicles) can
communicate with certain roadside infrastructures or base stations. Further,
the node mobility in a VANET is constrained to the road topologies,
whereas the movement of nodes in a MANET is more random in nature.
Considering that vehicles move over large distances at relatively high
speeds, a VANET undergoes fast topological changes. Another important
difference is that in a MANET, power is a major constraint but in VANET
the battery power available in a vehicle is quite adequate. The issues such
as the relatively larger size of VANETs compared to MANETs and the
relatively high speed with which vehicles move, need to be appropriately
considered for the design of an effective VANET.

7.11  Security Issues in a MANET

As discussed earlier, MANETS are fundamentally different from both

wired networks and infrastructure-based wireless networks. The nature of
MANETs not only introduces new security concerns but also exacerbates
the problem of detecting and preventing anomalous behaviour. In a wired
network or in an infrastructure-based wireless network, an intruder is
usually a host that is outside the network and therefore could be controlled
through a firewall and subjected to access control and authentication. In
a MANET, on the other hand, an intruder is part of the network, and
therefore much more difficult to detect and isolate.
Dynamic topological changes and the inherent wireless communications
in a MANET, make it vulnerable to different types of attacks. At the
physical layer, an intruder can easily cause jamming or overload the
available network resources beyond their capacities, thereby effectively
paralysing it. Wireless links can get jammed and the batteries at the nodes
can get depleted by such overloading, causing breakdowns of the network.
Attackers can also disturb the normal operation of routing protocols by
modifying the headers of packets. The intruder may insert spurious
 Mobile Ad Hoc Networks   139

information while routing packets, causing erroneous routing table updates

and thereby leading to frequent misroutings.
A few important characteristics of ad hoc networks that can be exploited
to cause security vulnerabilities are the following:

Lack of physical boundary:  Each mobile node functions as a router

and forwards packets from other nodes. As a result, network boundaries
become blurred. The distinction between nodes that are internal or external
to a network becomes meaningless, making it difficult to deploy firewalls
or monitor the incoming traffic.

Low power RF transmissions:  It is possible for a malicious node to

continuously transmit and monopolise the medium and cause its
neighbouring nodes to wait endlessly for transmitting their messages. Also,
signal jamming can lead to a denial-of-service (DoS) attack.

BOX 7.1  Security issues in MANETs

Simple attacks on a MANET may be in the form of broadcasting false
information or stealing information by eavesdropping. The injection of false
or malicious messages into a network can have serious repercussions. More
destructive attacks can be orchestrated by trojan horses and viruses bringing
down almost all the nodes in the network. There can be physical attacks as
well. These include wiretapping, theft of equipment, and storage media.
   In contrast to wired networks where the network devices are usually kept
behind locked doors, ad hoc network equipment is usually carried around as
small battery-powered devices. This makes them more vulnerable for attacks,
since they are easier to capture and carry away. Again it can be quite hard
to intercept wired media without getting noticed because the media itself
might be hard to get to and to intercept signals in the cables without drawing
attention is difficult. In the wireless medium, tapping is as easy as just putting
up an antenna, usually small enough to be noticed.

Limited computational capabilities:  Nodes in an ad hoc network usually

have limited computational capabilities. It therefore becomes difficult to
deploy compute-intensive security solutions such as setting up a public-
key cryptosystem. Inability to encrypt messages invites a host of security
attacks such as spoofing as well as several forms of routing attacks.

Limited power supply:  Since nodes normally rely on battery power,

an attacker might attempt to exhaust batteries by causing unnecessary
transmissions to take place or might cause excessive computations to be
carried out by the nodes.

Characteristics of secure ad hoc networks

We discuss below a few important characteristics of a secure ad hoc
network. Different types of attacks on the network attempt to breach one
140  Fundamentals of Mobile Computing

or more of these security features. A security solution should ensure that

these characteristics are not compromised. A secure ad hoc network should
have the following characteristics:

• Availability:  It should be able to survive denial-of-service (DoS)

attacks.
• Confidentiality:  It should protect confidentiality of information by
preventing its access by unauthorized users.
• Integrity:  It should guarantee that no transferred message has been
tampered with.
• Authentication:  It should help a node to obtain guarantee about the
true identity of a peer node.
• Non-repudiation:  It should ensure that a node having sent a message,
cannot deny it.

7.12  Attacks on Ad Hoc Networks

Out of the large varieties of attacks that are possible in a MANET, many
are essentially routing attacks. Several other attacks target the wireless
transmission in a MANET, especially the network connectivity. In this
section, we first discuss a classification of various types of MANET security
attacks and then discuss a few types of attacks that are rather common.
Various attacks on a MANET can be classified into passive and active
attacks, based on the means of attack. Passive attacks target to monitor and
steal the data exchanged in the network, without disrupting the network
operations. It becomes very difficult to identify these attacks since these
do not have any perceivable symptoms. These attacks can be reduced by
using suitable encryption techniques. An active attack, on the other hand, is
destructive and disturbs the normal functionality of the network. Table 7.2
shows a classification of various types of security attacks against MANETs
into active and passive types. We discuss these different types of attacks
in this section. Attacks can also be classified according to the specific layer
of the protocol stack that they target as shown in Table 7.3. Each attack
can be considered to be exploiting the vulnerabilities at one or more layers
of the MANET protocol stack while most attacks target certain security
vulnerabilities at specific protocol layers. The multilayer attacks are those
that exploit the vulnerabilities existing at more than one protocol layer. We
discuss a few of the important types of security attacks in the following.

TABLE 7.2  Passive and Active Attacks in a MANET

Passive attacks Active attacks

Snooping, eavesdropping, traffic Wormhole, black hole, grey hole, resource
analysis, monitoring consumption, routing attacks
 Mobile Ad Hoc Networks   141

TABLE 7.3  Attacks at Different Layers of a MANET Protocol Stack

Layer Attacks
Application layer Malicious code, repudiation, data corruption
Transport layer Session hijacking, SYN flooding
Network layer Wormhole, black hole, fabrication attack
Data link layer Resource consumption
Physical layer Traffic analysis, monitoring, disruption, jamming, inter-
ceptions, eavesdropping
Multilayer Denial-of-Service (DoS), impersonation, replay

Routing loop
By sending tampered routing packets, an attacker can create a routing loop.
This will result in data packets being sent around endlessly, consuming
bandwidth and causing dissipation of power for a number of nodes. In this
attack, the packets are prevented from reaching their intended recipients
and thus it can be considered to be a type of denial-of-service (DoS) attack.

Malicious code attacks

A malicious code can be a virus, worm, spyware, or a Trojan. In a MANET,
an attacker can propagate malicious code and can slow down the nodes,
overload the network, or even crash the nodes.

Repudiation attack
Repudiation attack refers to the denial of participation in a communication.
In this attack, a malicious user can deny a credit card or bank transaction.

SYN flooding attack

In this attack, an attacker creates a large number of half-opened TCT
connections with the victim nodes by sending a large number of SYN
packets to them. This causes the TCP connection tables of the victim nodes
to overflow.

Session hijacking
In a typical session, all the communications are authenticated only at the
beginning of the session. The attacker can spoof the IP address of a node
that has just started a session and hijack the session from the victim and
perform a DoS attack.

Fabrication attack
In AODV routing, when a node detects a broken link while forwarding a
packet (possibly because the next hop node has either moved or has shut
142  Fundamentals of Mobile Computing

down), it sends a route error message towards the packet sender. In the
fabrication attack, a malicious node sends a false route error message to
the packet sender, even when the next hop link is not broken.

Black hole
In this type of attack, a node can set up a route to some destination via
itself, and when the actual data packets are received from other nodes,
these are simply dropped. This node forms a black hole, to which data
packets enter but never leave.

Grey hole
A special case of the black hole attack is the grey hole attack. In this attack,
the attacker selectively drops some kinds of packets that pass through it
but not the others. For example, the attacker might forward routing packets
but not the data packets. This type of attack is more difficult to detect
compared to a black hole attack.

Partitioning
In this kind of attack, the attacker partitions a network by causing some
nodes to split up from the other nodes. That is, one set of nodes is not
able to communicate with the other set of nodes. By analysing the network
topology the attacker can choose to make the partitioning between the
set of nodes that causes the most harm to the system. This attack can be
accomplished in many ways, such as by tampering routing packets as in
the previous attacks. It can also be launched through some physical attack
such as radio jamming.

Blacklist
This attack tries to exploit a loophole in security mechanisms. Some ad
hoc routing protocols try to tackle this security problem by keeping a list
of perceived malicious nodes. Each node has a blacklist of, what it thinks,
bad nodes and thereby avoids using them when setting up routing paths.
An attacker might try to get a good node blacklisted, causing the other
good nodes to add this node to their respective blacklists and so avoid it.

Wormhole
In a wormhole attack, a direct link (tunnel) between the two nodes
is established. This is referred to as wormhole link. The direct link can
be established by making use of a wired line, a long-range wireless
transmission, or an optical link. Through the wormhole link, one node
eavesdrops messages at one end, and tunnells them through the wormhole
link to the other node which then replays them. The tunnel essentially
emulates a shorter route through the network and so naive nodes prefer
 Mobile Ad Hoc Networks   143

to use it rather than the alternative longer routes. Once a wormhole is

established, a malicious node can use it for traffic analysis or make a denial-
of-service attack by dropping certain data or control packets. When this
attack targets specifically the routing control packets, the nodes that are
close to the attackers are shielded from any alternative routes with more
than one or two hops to the remote location.

Dropping routing traffic

It is essential that in an ad hoc network, all nodes participate in the routing
process. However, it is possible that a node may act selfishly and process
only the routing information that is related to itself either maliciously or
to conserve energy. This behaviour/attack can create network instability
or can even segment the network.

7.13  Security Attack Countermeasures

Countermeasures against the various types of attacks we discussed, have
been proposed. We summarize the different security measures incorporated
in the different layers of the network protocol in Table 7.4. Though
cryptographic techniques are powerful techniques for ensuring confidentiality,
authentication, integrity, and non-repudiation, these are ineffective against
jamming. Spread spectrum technology such as frequency hopping is a
promising countermeasure against the signal jamming type of attacks. In
this technique, the transmission frequency changes randomly. Directional
antennae can be deployed as a countermeasure against signal jamming.

TABLE 7.4  Security Measures at Different Protocol Layers

Layer in protocol stack Security measures incorporated
Data link layer Use of spread spectrum transmission and directional
antennae
Network layer Use of authentication measures and keeping track
of the trusted nodes
Transport layer Securing and authenticating end-to-end commu-
nications through data encryption techniques
Application layer Detection and prevention of virus, worms, malicious
code through code analysis.

Maintenance of a trust rating of various nodes is a promising technique

to overcome many of the routing attacks. In this technique, every node
maintains a trust rating of various nodes and packet transmission is carried
out using a selective flooding technique.
144  Fundamentals of Mobile Computing

SUMMARY

Mobile ad hoc networks are becoming very popular. Their use in every
day applications is rapidly increasing. In this chapter, we briefly discussed
the important characteristics of ad hoc networks and compared them with
infrastructure-based wireless networks. We then discussed their various
applications, the challenges in designing an effective MANET, and the
importance of designing effective routing protocols in proper functioning
of an ad hoc network. We identified why the routing task in a MANET is
much more complex than that in a traditional network. We then discussed
a few important routing protocols that have been proposed for use in
MANETs. We discussed VANETs, as a special type of MANET, which are
rapidly emerging as popular networks having many applications. Though
MANETs are handy and appear to be a promising networking technology,
they suffer from many more security vulnerabilities than the infrastructure-
based networks do. To be practically useful, the security concerns for
these networks need to be satisfactorily understood and addressed. We
finally reviewed the security issues in MANETs and discussed the overall
approaches adopted by a few security solutions.

FURTHER READINGS

Abolhasan, M., T. Wysocki, and E. Dutkiewicz, “A review of routing

protocols for mobile ad hoc networks”, Ad Hoc Networks, Vol. 2,
pp. 1–22, 2004.
Aggelou, G. and R. Tafazolli, “RDMAR: A Bandwidth-efficient Routing
Protocol for Mobile Ad Hoc Networks”. In: ACM International Workshop
on Wireless Mobile Multimedia (WoWMoM), pp. 26–33, 1999.
Belding-Royer, E.M. and C. Perkins, “Evolution and future directions of the
ad hoc on-demand distance-vector routing protocol”, Ad Hoc Networks,
Vol. 1, pp. 125–150, 2003.
Haas, J., “A new routing protocol for the reconfigurable wireless networks”,
Proc. of IEEE Int. Conf. on Universal Personal Communications, pp. 562–566,
1997.
Jacquet, P., P. Muhlethaler, T. Clausen, A. Laouiti, A. Qayyum and
L. Viennot, “Optimized link state routing protocol for ad hoc networks”,
IEEE INMIC, Pakistan, 2001.
Jiang, M., J. Ji, Y.C. Tay, “Cluster based routing protocol”, Internet Draft,
draft-ietf-manet-cbrp-spec-01.txt , 1999.
Johnson, B. David and David A. Maltz, “Dynamic Source Routing in
Ad Hoc Wireless Networks”. In: Mobile Computing, T. Imielinski and
H. Korth, Eds., Kluwer Academics Publishers, pp. 153–181, 1996.
 Mobile Ad Hoc Networks   145

Johnson, D.B. and D.A. Maltz, “DSR: The Dynamic Source Routing Protocol
for Multihop Wireless Ad Hoc Networks”. In: Ad Hoc Networking,
C.E. Perkins, Ed., Addison Wesley, Boston, 2001.
Lee, S., W. Su, and M. Gerla. “On-demand Multicast Routing Protocol in
Multihop Wireless Mobile Networks”, Mobile Networks and Applications,
Vol. 7, Issue 6, December 2002.
Marina, M.K. and S.R. Das, “Routing Performance in the Presence of
Unidirectional Links in Multihop Wireless Networks”. In: Proc. of the
3rd Symposium of Mobile Ad Hoc Networking and Computing (MobiHoc),
Lausanne, Switzerland, June 2002.
McQuillan, John M., Isaac Richer, and Eric C. Rosen, “The New Routing
Algorithm for the ARPANet,” IEEE Trans. on Comm., 28(5), pp. 711–719,
1980.
Mohan, P. Madhan, Johnson J. James, Murugan K., and V. Ramachandran,
“A Comparative and Performance Study of On-demand Multicast
Routing Protocols for Ad Hoc Networks”, College of Engineering,
Guindy (CEG), Anna University.
Perkins, C.E. and T.J. Watson, “Highly dynamic destination sequenced
distance vector routing (DSDV) for mobile computers”. In: ACM
SIGCOMM’94 Conference on Communications Architectures, London, UK,
1994.
Perkins, E. Belding-Royer and S. Das, “Ad hoc On-demand Distance Vector
(AODV) Routing”, IETF Draft, November 2001.
Royer, E.M. and C.E. Perkins, “Multicast Operation of the Ad Hoc On-
demand Distance Vector Routing Protocol”, Proceeding of the fifth annual
ACM/IEEE international conference on mobile computing and networking,
pp. 207–218, 1999.
Siva Ram Murthy, C. and B.S. Manoj, Ad Hoc Wireless Networks: Architectures
and Protocols, Prentice Hall PTR, New Jersey, USA, May 2004.
Thomas Kunz and Ed Cheng, “On-demand Multicasting in Ad Hoc
Networks: Comparing AODV and ODMRP”, Carleton University.
Toh, C., “A novel distributed routing protocol to support ad hoc mobile
computing”. In: IEEE 15th Annual International Phoenix Conf., pp. 480–486,
1996.
Zhang, Yongguang, Wenke Lee, and Yi-an Huang, “Intrusion detection
for wireless ad hoc networks”, Wireless Networks 9, pp. 545–556, 2003.
Zhou, L. and Z. J. Haas, “Securing Ad Hoc Networks”, IEEE Network 13(6),
pp. 24–30, 1999.
146  Fundamentals of Mobile Computing

EXERCISES
1. What is an ad hoc network? Why the traditional routing strategies
cannot be deployed in a MANET straightaway? Compare the
MANET routing strategies with the routing strategies of traditional
networks.
2. Define the characteristics of MANETs that are important determinants
in the design of an effective ad hoc network.
3. What do you mean by dynamic topology of a MANET? What
problems does it cause in the design of a routing protocol? How are
these problems addressed in a popular MANET routing protocol?
4. Briefly explain why the traditional packet routing protocols for
wired networks cannot be used straightaway in a MANET. Discuss
how the routing protocols for traditional wired networks have been
extended to work in a MANET.
5. Describe at least three applications of mobile ad hoc networks.
6. What are the important design constraints on a MANET? Explain
the implications of these constraints.
7. Can cellular networks and wireless LANs be considered as ad hoc
networks? Explain your answer.
8. Discuss why routing is a more challenging problem in MANETs
when compared to wired networks. Explain the different types of
routing protocols in MANETs.
9. What is “counting-to-infinity” problem? How is this problem
addressed in a MANET?
10. What are the competing issues that are addressed by a routing
protocol in a MANET? How are these achieved?
11. Briefly explain the important classes of MANET routing protocols
and compare their relative advantages. Compare them with respect
to network overhead, routing quality and routing time.
12. What do you mean by unicast, multicast and broadcast
communications in a MANET? Give examples of typical applications
of each of these types of communication.
13. What is the difference between reactive and proactive routing in
MANETs? Give examples of each category of protocols.
14. What is a hybrid routing protocol? What is its advantage over the
other classes of routing protocols?
15. Explain the working of the Destination-Sequenced Distance-Vector
(DSDV) routing protocol using suitable examples.
16. Explain how the security threats in MANETs are different from a
wired network.
 Mobile Ad Hoc Networks   147

17. Explain Dynamic Source Routing (DSR). Clearly highlight the route
discovery and route maintenance operations.
18. What is a wireless sensor network? How is it different from a
MANET? Explain an application of a wireless sensor network.
19. Describe the security measures that can be incorporated in each layer
of protocol stack of MANETs.
20. Write short notes on:
(a)  Characteristics of a secure ad hoc network
(b)  Security attacks in ad hoc networks.
21. Discuss the important security threats in a MANET. What are the
factors responsible for limited security in MANETs?
22. Explain each of following types of attacks associated with ad hoc
networks:
(a) Routing loop   (b) Black hole   (c) Grey hole   (d) Partitioning
(e) Blacklist   (f) Wormhole   (g) Dropping routing traffic
23. Give an example of a core-based multicast routing protocol for
MANETs. Explain its working.
24. Mention the important differences between a mobile ad hoc network
and a cell phone network.
25. Explain any one routing technique that can be used in a mobile ad
hoc network.
26. What are the different categories of routing protocols for mobile ad
hoc networks?
27. Explain the factors that make mobile ad hoc networks more
vulnerable to security attacks compared to the traditional networks.
Also, explain the major types of security attacks that are possible in
a mobile ad hoc network. How can each of these types of attacks be
overcome?
28. What is the difference between active and passive security attacks
in a MANET? Give an example of each of these types of attacks.
29. What do you understand by active and passive attacks on a MANET?
Name at least five different types of security attacks in a MANET.
Classify these into active and passive attacks. Discuss the security
measures against these attacks.
30. For every layer of MANET protocol stack, discuss at least one type
of security attack that exploits a vulnerability at that layer.
31. A major task of the designer of a wireless sensor network is
prolonging the life of the network. Explain how this is achieved
while designing a MANET.
32. Explain why traditional IP-based routing protocols may not perform
satisfactorily in mobile ad hoc networks.
148  Fundamentals of Mobile Computing

33. What do you mean by size and node density of a MANET? Explain
these two terms and discuss how these two parameters impact the
design of a MANET.
34. What is a VANET? Explain how does it differ from a traditional
MANET? Explain any one application of a VANET.
35. Explain some routing attacks that are possible at the application
layer of a MANET.
 Wireless Sensor Networks
8
A wireless sensor network (WSN) is a self-organizing network of tiny
sensor nodes. Each sensor node (also called mote) usually senses certain
physical characteristics of its environment such as temperature, sound,
vibration, precipitation, etc. and then transmits the sensed information
to the user of the network. A WSN is called self-organizing since it gets
setup and configured automatically without requiring any form of manual
intervention.
A schematic diagram depicting the working of a WSN is shown in
Fig. 8.1. As can be observed from the figure, each node of a WSN senses
some specific data pertaining to the environment according to the user’s
specific requirements and the aggregate data is transmitted to the user of
the network. For transmission of the required data, the nodes set up an
ad hoc network among themselves.

Figure 8.1  A schematic model of a wireless sensor network.

If one analyses the emergence of WSNs, it can easily be observed that

the two most important contributory factors have been the recent progress
in wireless communication technology and digital electronics. While
advancements in wireless communications have made it possible for tiny
149
150  Fundamentals of Mobile Computing

sensor nodes to transmit data at low-power levels, advancements in digital

electronics have led to the development of sensor nodes that are small,
powerful and yet inexpensive. For example, a reasonably sophisticated
sensor node might cost less than rupees one hundred. The low cost of the
sensor nodes makes it possible to deploy them in large numbers in a WSN.

BOX 8.1  The mote

The name mote was given by a few scientists from the university of California
at Berkeley. These scientists were working on a project, called smart dust, which
was funded by the Defense Advanced Research Projects Agency’s (DARPA). The
aim of the Smart dust project was to shrink the size of the sensor nodes (motes)
to the size of dust. At this size of the motes, they can have very special uses.
For example, they can be ingested to help diagnose diseases. Since they can
float in air and liquids due to their small size, they can be used to monitor
inaccessible places such as estimating the size of an underground oil well or
the pollution level over a city.

WSNs are multi-hop wireless networks, similar to ad hoc networks.

However, these networks have sensor devices as nodes in contrast to hand-
held computers in ad hoc networks. Also, several thousands of nodes are
deployed in a WSN of size of a few hundred square metres, compared to
a few tens of nodes in the case of a comparable ad hoc network.
In the following section, we discuss the main differences between
WSNs and mobile ad hoc networks (MANETs).

8.1  WSN vs. MANET

Though there are significant differences between the two, WSNs can be
considered to be a special type of MANET. This is because WSNs are
ad hoc networks after all. Also, the network topology of both MANET
and WSNs changes dynamically. MANET topology changes due to user
movements, whereas the topology of a WSN changes continually as nodes
keep failing and node batteries keep getting depleted. In spite of these
superficial similarities, there are significant differences between these two
types of networks as enumered below.
1. Size and sophistication of the sensor nodes:  As already discussed in
Chapter 7, a MANET is usually made up of hand-held devices such
as laptops, palmtops, personal digital assistants (PDA) and other
such devices equipped with wireless transceivers. Compared to a
hand-held device, a sensor node is much more limited in power,
computational capacity, and memory. Therefore, sensor nodes
cannot perform complex computations such as encryption and run
 Wireless Sensor Networks   151

complex routing algorithms, which a hand-held device can. The

sensor nodes usually restrict themselves to sensing, aggregating, and
transmitting, and carrying out only very simple network routing
operations.
2. Node density:  The number of sensor nodes in a WSN is usually
several orders of magnitude higher than the nodes in an ad hoc
network of comparable size. The number of nodes in a WSN is
typically several thousands over an area of a few hundred square
metres compared to several tens of nodes in a MANET over a
comparable area. Thus, sensor nodes are much more densely
deployed in a WSN.
3. Node reliability:  Once the sensor nodes get deployed, it is hard
to access them for repair or replacement. Therefore, as the battery
power of a node gets depleted, it quickly becomes incapacitated.
The issue gets more complicated since these nodes are deployed in
the open and often under harsh weather conditions, thus increasing
their failure rates. Hence, sensor nodes are more prone to failures
than the hand-held devices. On the other hand, nodes in a MANET
do not fail as frequently since they have larger batteries and also
batteries are easily recharged. Of course, at present some types of
sensor nodes are able to recharge themselves from solar energy, from
vibrations, or from any existing temperature differentials.
4. Broadcast versus point-to-point communication:  Sensor nodes mainly
use broadcast communication, whereas most ad hoc networks are
based on some form of point-to-point communication.
5. Node identity:  Sensor nodes do not usually have global identification
(ID) because of the large number of sensors that are deployed.
Assigning and maintaining the global identifies in a dynamic network
would consume too much of computation and communicational
resources. Thus, it is very difficult for a user of WSN to address
or command sensor nodes either individually or in small groups.
MANET nodes, in contrast, do have IDs by which they can be
addressed and communicated to.
6. Node movement:  Nodes in a WSN are usually static, whereas nodes
in a MANET move as desired by the user.
7. Transmission range:  Sensor nodes have tiny batteries and therefore
can only afford to transmit at very small power levels. Also, they
are equipped with much simpler wireless transceiver equipment
compared to MANET nodes. As a result, the distance over which a
sensor node can directly communicate is many times smaller than
the distance over which a MANET node can communicate.
8. Bandwidth:  The bandwidth available to a node in a WSN is typically
of the order of 10 kbps, whereas the nodes in a MANET typically
152  Fundamentals of Mobile Computing

transmit at much higher rates. This is because the sensor nodes do

not have a DSP processor and as a result cannot perform digital
signal modulation.
9. Communication protocol:  The communication protocols used in
MANETs incur too much computing and transmission overheads
to be meaningfully deployed in a WSN. WSNs deploy much simpler
protocols.
10. Information flow:  The information flow in a WSN is typically well-
structured and the flow is in one direction only. This is so because
the nodes essentially collect information from the environment and
transmit it to a server. This characteristic has been exploited to
develop simpler and efficient communication protocols for WSNs. On
the other hand, in a mobile ad hoc network, the flow of information
is usually chaotic.

8.2  Applications

WSNs were conceived about two decades ago essentially for military
applications such as battlefield surveillance and enemy tracking. But,
nowadays WSNs have become popular in several civilian areas such
as habitat monitoring (Cerpa et al., 2001 and Mainwaring et al., 2002),
environment observation and forecasting (Biagioni and Bridges, 2002),
health applications (Akyildiz et al., 2002), and home and office applications
(Srivastava et al., 2001). The following are a few important applications of
WSNs:

Military applications
WSNs have been used to provide the following services in battlefields:
Monitoring enemy troop formations:  Wireless sensor networks can be
used by a military commander to gather data about enemy formations. The
gathered data could include troop strength, location of troops, amount and
type of war equipment, etc. Such collected information can be used by the
commander to decide about counter deployments.
Battlefield surveillance:  To monitor enemy intrusions or to detect the
presence of enemy troops, sensors can be randomly deployed in inaccessible
and critical regions. This way, intrusions and army advancements can be
easily determined. Based on the surveillance data, safe routes or paths for
the offensive attacks can be worked out.

Great Duck Island (GDI) system

This is an example of WSN deployment in a habitat monitoring application.
The researchers from Intel Research laboratory deployed a mote-based
 Wireless Sensor Networks   153

sensor network in Great Duck Island, Maine, USA, to monitor the behaviour
of storm petrel (small sea birds). The important features of their designed
network were as follows:
Internet connectivity:  The sensor network at Great Duck Island (GDI)
was designed to be accessed via the Internet. It was also possible to monitor
and manage the sensor networks deployed in the remote sites over the
Internet.
Tree topology:  The nodes were arranged in clusters and the clusters
organized in a tree topology. To start with, three to four clusters of 100
fixed sensor nodes were deployed.
Life of the sensor network:  The sensor network was designed to function
for at least a couple of months without the need for recharging. However,
GDI had sufficient solar power to recharge the sensor nodes that were
fitted with small solar panels.

PODS—A remote ecological sensor network

PODS was a research project in the University of Hawaii that built a
wireless sensor network to investigate as to why endangered species of
plants grow in one area but not in neighbouring areas. They deployed
secret sensor nodes, called Pods, in Hawaii Volcanoes National Park. Each
pod consisted of a computer, radio transceiver and environmental sensors
including a high resolution digital camera. These pods relayed sensor data
to the user via a wireless link to the Internet.

Automated Local Evaluation in Real Time (ALERT)

ALERT is an automated flood alarm system, probably the first well-known
deployment of a real wireless sensor network. ALERT provides important
real-time rainfall and water level information to evaluate the possibility
of potential flooding. ALERT sensor networks are usually equipped with
meteorological sensors, such as water level sensors, temperature sensors,
and wind sensors. Data is transmitted from the sensor site to the base
station. A flood forecast model is deployed to process the received data
and issue automatic warnings. Currently, ALERT is deployed across most
of the states of western United States such as California and Arizona.

Smart Sensors and Integrated Microsystems (SSIM)

SSIM is a biomedical project about the artificial retina. In this, a retina
prosthesis chip (prosthesis is an artificial replacement of a missing body
part) consisting of 100 micro sensors is built and implanted within the
human eye. This allows a patient with no vision or limited vision to see at
an acceptable level. Feedback control, image identification and validation
are achieved using wireless communication with a processor.
154  Fundamentals of Mobile Computing

8.3  Architecture of the Sensor Node

A typical architecture of a sensor node is shown in Fig. 8.2. As can be

seen, the sensor nodes pass information to the controller through an
ADC (analog-to-digital signal converter). The controller processes this
information and often filters it before transmitting it with the help of the
transmitter.
In the following, we briefly explain the different components of the
sensor node architecture.

Controller
The controller is possibly the most important component of the sensor
node. Besides responding to certain user commands, it processes the
sensor information, transmits it on the network, and also performs
several network-related operations. A typical controller consists of an
8-bit 4-MHz microcontroller with slow 10-Kbps communication capability
and 8-Kbyte read-only memory and a 512-byte RAM. While a controller
is commonly designed using a microcontroller, SoC (System on Chip) and
ASIC (Application Specific Integrated Circuit) are the other alternatives.

Transceiver Sensor

Controller ADC Sensor

Battery

External Memory Sensor

Figure 8.2  Architecture of a sensor node.

Transceiver
In a transceiver, the functionality of both the transmitter and receiver are
combined into a single device. Most WSNs are based on radio frequency
communications among the sensor nodes. WSNs typically use the
licence-free communication frequencies such as the 915-MHz band or the
2.4-GHz band. Even when idle, most sensors consume power at a rate
that is almost equal to the power consumed while receiving signals. This
is because citcuits such as the amplifier and filter continue to operate as
usual. Consequently, when a sensor node is not transmitting or receiving,
it is better to completely shut down the transceiver rather than leave it in
the idle mode.
 Wireless Sensor Networks   155

External memory
From an energy perspective, the most appropriate memory is the on-chip
memory of a microcontroller or SoC. For storage of large amounts of data,
flash memories are used due to their low cost and high storage capacity.
Magnetic memories are rarely used due to their high energy consumption,
volume, and weight.

Power source
A sensor node consumes power for sensing, communicating and data
processing. Typically, more energy is required for data communication
than for any other process. Current sensors are able to renew their energy
from solar radiation or from vibrations. Two power saving policies are
being used. These are the Dynamic Power Management (DPM) policy and
the Dynamic Voltage Scaling (DVS) policy. DPM helps conserve power by
shutting down those parts of the sensor node that are idle. The DVS scheme
is based on varying the power levels based on the work load.

Sensors
Sensors are designed to exploit certain physical principles to produce
a measurable response to a change in some physical parameter such as
temperature or pressure. A sensor should be small in size and consume
negligible energy. Sensors fall into different categories based on whether
they are passive or active or omni-directional or narrow-beam sensors.
Passive sensors sense data without disturbing the environment in any
way and consequently consume almost no energy. Active sensors probe by
perturbing the environment in some way. Examples of active sensors are
sonar, ultraviolet devices, and radar. Naturally, these sensors need battery
power for their operation. Narrow-beam sensors sense in a well-defined
direction such as a camera. Omni-directional sensors cannot be ascribed
any direction to their measurements.

8.4  Challenges in the Design of an Effective WSN

Effective design of WSNs must take into account various constraints and
limitations that these networks have to work with. Besides the limited
processing capability and available memory, these networks work with
limited bandwidth and limited network coverage. Further, new application
scenarios lead to new types of challenges. The following subsections
identify areas that constitute some of the important challenges in the
design of an effective WSN.

Energy:  Sensor nodes are powered by batteries and it is usually difficult to

replace them when depleted since the sensor nodes are usually deployed in
156  Fundamentals of Mobile Computing

remote and hostile environments. So the hardware, the network protocols

and the applications must be designed to minimize energy consumption
so as to increase network life.

Location discovery:  In many applications, routing protocols need the

geographic location of each sensor node to forward data among the
networks. Location discovery protocols must be designed in such a way that
location discovery can be done accurately and with minimum overhead.
Use of WSNs becomes too expensive for applications.

Cost:  Cost is another important factor that needs to be optimized during

the design of WSNs. If the initial cost is high, popularization of WSNs
would be difficult.

Security:  An important application of sensor networks is the deployment

in mission-critical tasks such as military surveillance, security, etc.
However, the secure operation of sensor networks is even more difficult
compared to MANETs due to the severe resource limitations of sensor
nodes. Security solutions are constrained when applying them to sensor
networks. For example, the encryption of transmitted data incurs significant
processing overhead and power consumption. Further, sensor nodes are
more susceptible to tampering and capture since they function totally
unattended. Some of the important issues that need to be addressed in a
security context are: secure routing, authentication, key establishment and
trust setup, and prevention of physical attacks against sensor nodes.

8.5  Characteristics of Sensor Networks

The following subsections identify some important characteristics of sensor

networks.

Multi-hop wireless communication:  Due to low energy and simple

antennae, the wireless coverage of sensor nodes is extremely small.
Therefore, multi-hop communication is the norm in these networks.

Energy efficient operation:  A key objective of WSNs is to maximize the

life of the network.

Self-organization:  This is an important characteristic of WSNs and helps it

to cope with node failures. Failures may occur due to exhaustion of battery
power and several other reasons such as hardware or software failure. The
appearance of obstacles in communication paths and the addition of nodes
to the network would also require self configuration.

Collaboration and in-network processing:  Depending on the application,

it is often required that a group of sensor nodes interact among themselves
to detect an event by meaningfully processing their collective information.
 Wireless Sensor Networks   157

Data redundancy:  In a sensor network, it is not possible to assign a

global identifier to each node due to the sheer number of nodes deployed.
Lack of such global identifiers along with random deployment of sensor
nodes, makes it hard to select a specific set of sensor nodes to be queried.
Therefore, it is very difficult to give a command for obtaining data from the
sensor nodes that are spread over a specific area or are part of the network.
For this reason, data is usually transmitted from every sensor node within
the deployment region with significant redundancy.

8.6  WSN Routing Protocols

In a wireless sensor network, nodes do not have any global identification

associated with them. Further, nodes fail frequently due to a variety of
reasons leading to dynamic changes to the network topology. As nodes
fail, some network links get destroyed while some new links may get
established when a new sensor node is introduced. It is easy to see that
such an environment cannot be served efficiently by routing protocols
developed for wired networks. It also needs to be realized that the network
protocols for MANET are not suitable due to their limited computing
power, lack of a DSP processor, limited battery energy, and the simplicity
of antenna of the sensor nodes. In the following subsection, we classify the
large number of WSN routing protocols that have been suggested based
on the protocol used and the network structure.

8.6.1  Classification Based on Protocol Operation

In WSNs, there are essentially four main types of protocols that have been
proposed. These protocols are briefly discussed below.

Multipath-based routing:  In a network such as WSN, where the topology

changes dynamically, established paths dissolve sometimes even before the
full message transmission is complete. Since a path setup does take time
and is expensive, it is necessary to set up redundant paths and maintain
them as backups. Multipath-based routing protocols have at least one
alternative path (from source to destination). However, this approach
increases energy consumption and traffic. The redundant paths are kept
alive by sending periodic messages. An example of such a protocol is
Directed Diffusion (DD) paradigm (Intanagonwiwat et al., 2000 and Estrin
et al., 1999).

Query-based routing:  In these protocols, the destination nodes query data

through the network. The node(s) containing the required data transmit it
back to the node that has initiated the query along that path. An example
of a protocol deploying this technique is Rumor routing (Braginsky and
Estrin, 2002).
158  Fundamentals of Mobile Computing

Negotiation-based routing:  This class of routers use negotiation as the

basis for cooperation between the competing entities. Inter domain routing
is often driven by self-interest and is based on a limited view of the internet,
which may be detrimental to the stability and efficiency of routing. An
example of this class of routers is Sensor Protocols for Information via
Negotiation (SPIN) (Heinzelman et al., 1999).

QoS-based routing:  In some applications, data is required to be delivered

within a certain period of time from the moment it is sensed, otherwise
it would become stale and be of little use. Therefore, bounded latency for
data delivery is a condition for time-constrained applications. In WSNs,
however, the conservation of energy is often considered relatively more
important than the quality of data sent. As the energy gets depleted, the
network may be required to reduce the quality of the results in order to
reduce the energy dissipation in the nodes and hence lengthen the total
network lifetime. An example of a QoS-based routing protocol is Sequential
Assignment Routing (SAR) (Akyildiz et al., 2002 and Sohrabi et al., 2000).

8.6.2  Classification Based on Network Structure

In general, routing techniques deployed in a WSN can be classified, based

on the network structure, into either flat routing, hierarchical routing, or
location-based routing.

Flat routing:  In this type of routing, each node performs routing. An

example of flat routing is the gradient-based routing technique (Schurgers
and Srivastava, 2001).

Hierarchical routing:  These protocols are also known as cluster-based

routing. In these protocols, different nodes can play different roles in the
network based on their remaining battery powers. Nodes having higher
battery energy are used to process and send the information, while low-
energy nodes only focus on sensing activity. Examples of such protocols
include Low Energy Adaptive Clustering Hierarchy (LEACH) (Heinzelman
et al., 2000), Power-Efficient Gathering in Sensor Information Systems
(PEGASIS) (Lindsey and Raghavendra, 2002 and Lindsey et al., 2001).

Location-based routing:  In these protocols, the nodes are addressed

by their location. Distances to neighbouring nodes can be estimated by
the received signal strengths (RSS) or by GPS receivers. In this type of
protocol, the positions of sensor nodes are exploited to route data to the
network. The distance between the neighbouring nodes can be estimated
on the basis of the incoming signal strength. The relative coordinates of
neighbouring nodes can be obtained by exchanging location information
between neighbours. If there is no activity, nodes go to sleep to save energy.
To maximize energy saving, it is necessary to have as many sleeping nodes
 Wireless Sensor Networks   159

in the network as possible. Examples of such protocols include Geographic

Adaptive Fidelity (GAF) (Xu et al., 2001) and Geographic and Energy
Aware Routing (GEAR) (Yu et al., 2001).
In the following subsections, we briefly review a few important
protocols belonging to the above three categories.

8.6.3  Directed Diffusion (DD)

In this protocol (Intanagonwiwat et al., 2000 and Estrin et al., 1999), each
sensor node tags its data with one or more attributes. A node interested
in some data floods the network with an interest packet. Interests are
flooded over the network. When a node receives an interest packet from
a neighbour, it responds by sending data to the neighbour to retrace the
path of the interest packet. Each node only knows the neighbour from
whom it got the interest. It is possible that each node would receive
the same interest from more than one neighbour. In this case, multiple
paths can be set up from the source node to the destination node. Among
these paths, one or a few high rate paths are defined and the other paths
remain low rate. Depending on the number of paths that are reinforced,
directed diffusion can be single-path or multi-path routing. If a better
path emerges, the protocol updates the paths. Further, directed diffusion
allows intermediate nodes to cache and aggregate data to improve data
accessibility and energy efficiency.

8.6.4  Rumor Routing

In rumor routing (Braginsky and Estrin, 2002), each node maintains a

neighbour list and an event table. When a node witnesses an event, it adds
the event to its event table. Nodes that have recently observed an event,
generate an agent with a certain probability. An agent is a long-lived packet
that roams through the network, propagating information. Each agent
carries a list of events it has encountered, along with the number of hops
to that event. When it arrives at a node, it synchronizes its list with the
node’s list. The agent traverses the network for some number of hops, and
then dies. Any node can generate a query that is destined to a particular
event. If it has a route to the event, then it will transmit the query. If it
does not, it will forward the query to a random neighbour. If the query
has not reached the destination node when the query Time-to-Live (TTL)
expires, then it will retransmit the query, or flood the query. The goal of
rumor routing is to avoid expensive flooding operations. Unlike directed
diffusion (DD), which tries to find an optimal path by flooding queries
for gradient setup, rumor routing sends a query on a random walk until
it finds the event path.
160  Fundamentals of Mobile Computing

8.6.5  Sequential Assignment Routing (SAR)

Sequential Assignment Routing (SAR) (Akyildiz et al., 2002 and Sohrabi

et al., 2000) uses the QoS information in its routing decisions. This protocol
creates trees rooted at one-hop neighbours of the sink by taking the QoS
metric, the available energy on each path and the priority level of each
packet into consideration. By using created trees, multiple paths from
sink to sensors are formed. One of these paths is selected according to the
available energy resources and QoS on the path.

8.6.6  Low Energy Adaptive Clustering Hierarchy (LEACH)

Low Energy Adaptive Clustering Hierarchy (LEACH) (Heinzelman et al.,

2000) was proposed by Heinzelman. It is a very popular routing protocol
for WSNs that is based on setting up a hierarchical structure. This protocol
divides the network into clusters (sections). In each cluster, an elected
sensor node acts as the cluster head. The task of a cluster head is to manage
communication among other cluster heads. A cluster head can process the
collected data before relaying to the other cluster heads. The cluster heads
collaborate to send data to the base station.

8.6.7  Power-Efficient Gathering in Sensor Information Systems

(PEGASIS)

In PEGASIS (Lindsey and Raghavendra, 2002 and Lindsey et al., 2001),

each node communicates only with the immediate neighbour. When a
node on the chain receives data from a neighbour, it aggregates the data
with its own data and sends the data to the next neighbour on the chain.
Rather than multiple cluster-heads sending data to the base station as in
LEACH, only one node on the chain is selected to transmit the data to
the base station.

8.6.8  Geographic and Energy Aware Routing (GEAR)

This protocol is based on the knowledge of a node’s own as well as its
neighbours’ locations and energy information (Estrin and Govindan, 2001).
It conserves energy by minimizing the number of interests in directed
diffusion (DD) by only considering certain regions rather than sending the
interests to the whole network.

8.6.9  Geographic Adaptive Fidelity (GAF)

The Geographic Adaptive Fidelity (GAF) (Xu et al., 2001) is an energy-
aware location-based routing algorithm. Though this class of protocols is
 Wireless Sensor Networks   161

designed primarily for mobile ad hoc networks, it has been used for sensor
networks as well. The network area is first divided into fixed zones forming
a virtual grid. Inside each zone, nodes collaborate with each other to play
different roles. For example, nodes elect one sensor node to stay awake for
a certain period of time while the others go to sleep mode.

8.7  Target Coverage

The coverage problem is one of the active issues of the WSNs that
determines how efficiently the sensor network is being covered by a set
of sensor nodes. This problem deals with the QoS of the network ensuring
that the particular sensor network is monitored or observed by at least
one sensor node. It may be broadly classified into the following categories:
• Area coverage where the sensor nodes are deployed.
• Target coverage where the sensor nodes are deployed to cover a
specific set of targets or points.
In the target coverage problem, a fixed number of targets (t1, t2, ,
tm) need to be continuously monitored by a number of sensor nodes (s1,
s2, , sn), while the lifetime of the network needs to be maximized. There
are a specific number of targets which are required to be covered by a set
of sensor nodes. Each target needs to be monitored by at least one sensor
node. Since sensor nodes are provided with only limited resources and
cannot withstand extreme environmental conditions, they are deployed in
large numbers for exceeding the actual requirements.
Some of the possible strategies to cover a specific set of targets are
described below.

8.7.1  Some Strategies for Target Coverage

Activating all the sensors at a time

In order to cover all the specific targets, the simplest method is to activate
all the sensor nodes deployed for coverage at the same time. However,
this simultaneous activation of sensor nodes may make them exhaust their
energy together. So, the lifetime of the network is the least in this scheme.

Formation of disjoint set covers of sensor nodes

An energy efficient method to cover targets is to make the sensor nodes
alternate between active and sleep modes (Cardei and Du, 2005). Of the
different sensors covering a target, some go to sleep whereas the remaining
stay active. This way there is no disruption of the monitoring process,
whereas different nodes get sleep at different times. During the active
162  Fundamentals of Mobile Computing

mode, a sensor node is capable of keeping track of a target and collect

important information pertaining to it. Disjoint set covers of active sensor
nodes are made in such a manner that each set cover covers all the targets.
These set covers are activated one after another till the sensor nodes are out
of their energy. So, the main idea behind this protocol is that rather than
making all the sensor nodes active at a time, these can be partitioned into
a number of disjoint set covers with each set cover capable of monitoring
the targets.

Formation of non-disjoint set covers of sensor nodes

Another energy efficient method for target coverage (Cardei et al., 2005) is
to make the sensor nodes part of more than one set cover, with the sensor
nodes alternating between the active and sleep modes. Several non-disjoint
set covers of active sensor nodes are made to activate successively where
each set cover is capable of keeping track of all the specific targets until
the time of energy exhaustion of sensor nodes. It results in much more
energy efficient monitoring compared to the disjoint set cover method of
target coverage, thus maximizing the network lifetime to a greater extent.

Partial target coverage

Till now, we have mentioned full coverage of targets which ensures that
all the required targets are to be covered. Information collected by a subset
of targets may also be beneficial (Zorbas et al., 2009). A number of set
covers of sensor nodes may be activated for partial coverage of the targets,
not necessarily covering all the targets. The gathered information about a
subset of targets by the sensor nodes may prove to be very important where
the main objective is to maximize the lifetime of the network.

K-coverage
In the k-coverage (k  1) approach (Zorbas et al., 2009), a minimal subset of
the sensor nodes is determined that can keep track of the required targets,
with each target being covered by at least k sensor nodes. The activated
nodes collectively gather information about the specific targets and send
it to the base station.

SUMMARY

In this chapter, we discussed sensor networks and the challenges in their

design and operation. We also reviewed the routing protocols used in
WSNs. A wireless sensor network consists of a large number of tiny
autonomous, lightweight and low cost devices called sensor nodes. A sensor
node has characteristics such as limited sensing and limited computations
 Wireless Sensor Networks   163

and hence can communicate only within short distances. The sensor nodes
are constrained in terms of energy, memory, computational capability and
communication bandwidth.
These sensors monitor different conditions at different locations, such
as temperature, humidity, lighting condition, pressure, soil makeup, noise
levels and help to find some current characteristics of certain objects such
as the presence or absence, mechanical stress, speed, direction, size, etc.
WSNs are practically becoming important. Various areas of applications
of sensor networks include critical missions, civil, healthcare, and
environmental disciplines, etc. The low cost of these nodes helps in their
deployment ranging from thousands to million nodes. These nodes can be
deployed either in a random fashion or in a pre-designed fashion.

FURTHER READINGS

Akyildiz, I.F., et al., “Wireless sensor networks: a survey”, Computer

Networks, Vol. 38, pp. 393–422, March 2002.
Akyildiz, I.F., W. Su, Y. Sankarasubramaniam, and E. Cayirci, “A survey on
sensor networks”, IEEE Communications Magazine, 40(8), pp. 102–114,
August 2002.
Biagioni, Edoardo and Kent Bridges, “The application of remote sensor
technology to assist the recovery of rare and endangered species”. In:
Special issue on Distributed Sensor Networks for the International Journal of
High Performance Computing Applications, Vol. 16, No. 3, August 2002.
Braginsky, D. and D. Estrin, “Rumor Routing Algorithm for Sensor
Networks”. In: Proceedings of the First Workshop on Sensor Networks and
Applications (WSNA), Atlanta, GA, October 2002.
Cardei, M. and D.Z. Du, “Improving Wireless Sensor Network Lifetime
through Power Aware Organization”, ACM Wireless Networks, Vol. 11,
No. 3, pp. 333–340, May 2005.
Cardei, M., M.T. Thai, Y. Li, and W. Wu, “Energy-efficient target coverage
in wireless sensor networks,” Proceedings of IEEE INFOCOM’05, Miami,
FL, pp. 1976–1983, 2005.
Cerpa, A., J. Elson, D. Estrin, L. Girod, M. Hamilton, and J. Zhao,
“Habitat monitoring: Application driver for wireless communications
technology”. In: Proceedings of the 2001ACM SIGCOMM Workshop on
Data Communications in Latin America and the Caribbean, April 2001,
2001.
Estrin, D., et al., “Next Century Challenges: Scalable Coordination in Sensor
Networks”. In: Proceedings of the 5th annual ACM/IEEE international
conference on Mobile Computing and Networking (MobiCom’99), Seattle,
WA, August 1999.
164  Fundamentals of Mobile Computing

Heinzelman, W., A. Chandrakasan, and H. Balakrishnan, “Energy-efficient

communication protocol for wireless sensor networks”. In: Proceedings
of the Hawaii International Conference System Sciences, Hawaii, January
2000.
Heinzelman, W., J. Kulik, and H. Balakrishnan, “Adaptive protocols for
information dissemination in wireless sensor networks”. In: Proceedings
of the 5th Annual ACM/IEEE International Conference on Mobile Computing
and Networking (MobiCom’99), Seattle, WA, August 1999.
Intanagonwiwat, C., R. Govindan, and D. Estrin, “Directed diffusion: A
scalable and robust communication paradigm for sensor networks”.
In: Proceedings of the 6th Annual ACM/IEEE. International Conference on
Mobile Computing and Networking (MobiCom’00), Boston, MA, August
2000.
Lindsey, S. and C.S. Raghavendra, “PEGASIS: Power Efficient Gathering
in Sensor Information Systems”. In: Proceedings of the IEEE Aerospace
Conference, Big Sky, Montana, March 2002.
Lindsey, S., C.S. Raghavendra and K. Sivalingam, “Data Gathering in
Sensor Networks using the Energy*Delay Metric”. In: Proceedings of the
IPDPS Workshop on Issues in Wireless Networks and Mobile Computing,
San Francisco, CA, April 2001.
Mainwaring, Alan, Joseph Polastre, Robert Szewczyk, David Culler, and
John Anderson, “Wireless sensor networks for habitat monitoring”. In:
ACM International Workshop on Wireless Sensor Networks and Applications
(WSNA’02), Atlanta, GA, September 2002.
Schurgers, C. and M.B. Srivastava, “Energy efficient routing in wireless
sensor networks”. In: MILCOM Proceedings on Communications for
Network-Centric Operations: Creating the Information Force, McLean, VA,
2001.
Schwiebert, Loren, K. Sandeep, S. Gupta, and Jennifer Weinmann, “Research
challenges in wireless networks of biomedical sensors”. In: Mobile
Computing and Networking, pp. 151–165, 2001.
Sohrabi, K., et al., “Protocols for self-organization of a wireless sensor
network,” IEEE Personal Communications, Vol. 7, No. 5, pp. 16–27,
October 2000.
Srivastava B., Mani, Richard R. Muntz, and Miodrag Potkonjak, “Smart
kindergarten: sensor-based wireless networks for smart developmental
problem-solving environments”. In: Mobile Computing and Networking,
pp. 132–138, 2001.
Xu, Y., J. Heidemann, and D. Estrin, “Geography-informed energy
conservation for ad hoc routing”. In: Proceedings of the 7th Annual
ACM/IEEE International Conference on Mobile Computing and Networking
(MobiCom’01), Rome, Italy, July 2001.
 Wireless Sensor Networks   165

Yu, Y., D. Estrin, and R. Govindan, “Geographical and Energy-Aware

Routing: A Recursive Data Dissemination Protocol for Wireless Sensor
Networks,” UCLA Computer Science Department Technical Report, UCLA-
CSD TR-01-0023, May 2001.
Zhou, Zongheng, Samir R. Das, and Himanshu Gupta, “Connected
k-coverage problem in sensor networks”, Proceedings of the International
Conference On Computer Communications and Networks (ICCCN 2004),
October 11–13, 2004, Chicago, IL, USA 2004.
Zorbas, D., D. Glynos, and C. Douligeris, “Connected partial target coverage
and network lifetime in wireless sensor networks”, Wireless Days (WD),
2009 2nd IFIP, pp. 1–5.

EXERCISES
1. Compare WSNs with MANETs with respect to their characteristics,
applications and challenges to their commercialization?
2. Discuss a classification of the routing protocols for sensor networks.
3. Write short notes on each of the following:
(a) Multipath-based Routing   (b) QoS-based Routing  
(c) Hierarchical Routing   (d) Location-based Routing
4. What is flat routing? How is it different from hierarchical routing?
5. Explain hierarchical routing and query-based routing in a wireless
sensor network.
6. What is a wireless sensor network (WSN)? Discuss some important
applications of WSNs.
7. Discuss the architecture of a sensor node in a wireless sensor network.
8. Explain the similarities and differences between a wireless sensor
network and mobile ad hoc network.
9. Explain the constraints and limitations on the operation and design
of WSNs.
10. Explain the main differences between the routing protocols for
WSNs and MANETs. Compare and contrast them with the routing
protocols for traditional wired networks. Discuss the main reasons
behind the differences in the routing approaches.
11. Why is design of a WSN considered to be more challenging than
that of a MANET? Explain your answer.
12. Explain the alternative WSN designs that can be proposed for
monitoring a finite number of discrete targets.
13. Do you agree with the statement: “Wireless sensor networks are a
special type of mobile ad hoc networks.” Justify your answer.
 Operating Systems for
9
Mobile Computing

O f late, smartphones are being used not only to make phone calls, but
also to carry out a host of other useful operations such as to make
video conference calls, send multimedia messages, take pictures, play
media files, browse World Wide Web (WWW), run remote applications,
etc. This sophistication of the present-day smartphones enables multiple
tasks to be run on the device. Consequently, a powerful operating system
has become an essential part of every smartphone. However, we need to
keep in mind that the capability of the operating system that is used in
a mobile handset varies a great deal with the degree of sophistication of
the device. In a mobile handset, the operating system (OS) performs two
main responsibilities.

9.1  Operating System Responsibilites in Mobile Devices

9.1.1  Managing Resources

An important responsibility of the operating system of a mobile device is

to facilitate efficient utilization of the resources of the device by performing
multiple tasks. The resources that are managed by the operating system
include processor, memory, files, and various types of attached devices
such as camera, speaker, keyboard, and screen. Typically, a mobile device is
expected to run multiple applications at the same time and each application
may in turn require running multiple tasks. A task can have multiple
threads. A few examples of such applications include voice communication,
text messaging, e-mail, video play, music play, recording, web browsing,
running remote applications, etc. As an example scenario of usage of a
smartphone, consider the following: a person might be listening to music,
at the same time he might answer an incoming call, and an SMS might
166
 Operating Systems for Mobile Computing   167

arrive at the same time which he might like to look-up while the call is still
on. Such a scenario requires concurrent execution of multiple tasks. When
multiple tasks contend to use the same set of resources, the OS acts like a
traffic cop—ensuring that different tasks do not interfere with each other.

9.1.2  Providing Different Interfaces

The operating system of a mobile device on the one hand provides a highly
interactive interface to the user of the device and on the other interfaces
with other devices and networks. An important interface concerns control,
data, and voice communications with the base station using different
types of protocols. Besides, an OS takes care of recognizing inputs from
the keyboard, sending outputs to the display screen, and interfacing with
peripheral devices such as other mobile devices, computers, printers, etc.
For the sake of brevity, we shall refer to the operating system used
in a mobile hand-held device as a mobile OS. At present, the mobile OS
marketplace is dominated by Symbian, Android, Windows mobile, Palm
OS, iOS, and Blackberry OS. In this chapter, after discussing a few basic
concepts relevant to mobile OS and the special constraints under which
a mobile OS is expected to operate, we will survey the evolution and
important features provided by a few popular mobile operating systems.
Finally, we will provide a brief overview of the operating systems designed
for the sensor nodes in a WSN.

9.2  Mobile O/S—A Few Basic Concepts

Traditionally, the operating system is viewed as providing a set of services

to the application programs. The operating system is usually structured
into a kernel layer and a shell layer. The shell essentially provides facilities
for user interaction with the kernel. The kernel executes in the supervisor
mode and can run privileged instructions that could not be run in the user
mode. During booting, the kernel gets loaded first and continues to remain
in the main memory of the device. This implies that in a virtual memory
system, paging does not apply to the kernel code and kernel data. For this
reason, the kernel is called the memory resident part of an operating system.
The shell programs are usually not memory resident. The kernel of the
operating system is responsible for interrupt servicing and management
of processes, memory, and files.
The traditional operating systems such as Unix and Windows are
known to have a monolithic kernel design. In a monolithic kernel OS design,
the kernel essentially constitutes the entire operating system code, except
for the code for the shell. The principal motivation behind this monolithic
design was the belief that in the supervisor mode, the operating system
168  Fundamentals of Mobile Computing

services can run more securely and efficiently. On the other hand, the
main problem with the monolithic kernel design is that it makes the kernel
massive, non-modular, hard to tailor, maintain, extend, and configure.
Considering the disadvantages of the monolithic kernel design, the
microkernel design approach has been proposed. The microkernel design
approach tries to minimize the size of the kernel code. Only the basic
hardware-dependent functionalities and a few critical functionalities
are implemented in the kernel mode and all other functionalities are
implemented in the user mode. Most of the operating system services run
as user level processes. The main advantage of this approach is that it
becomes easier to port, extend, and maintain the operating system code. The
kernel code is very difficult to debug compared to application programs.

Client Memory Network

App Server Server

Process File Display

Server Server Server

User Mode

Kernel Mode
Request
Microkernel
Reply

Hardware

(b) Microkernel

Figure 9.1  Monolithic design versus microkernel design of an operating system.

 Operating Systems for Mobile Computing   169

The reason for this is that a bug in a kernel code can crash the system,
thus crashing the debugger too. Further, even when some operating system
service crashes while being used by a user, it does not bring down the
entire system. This is one reason as to why a microkernel operating system
could be expected to be more reliable than an equivalent monolithic kernel
operating system. The overall architectural difference between a monolithic
kernel and a microkernel architecture is schematically shown in Figure 9.1.
To restrict the size of the kernel of a mobile OS to the minimum, most
mobile OS are, to different extents, based on the microkernel design.

9.3  Special Constraints and Requirements of Mobile O/S

There are a few special constraints under which the operating system of
a mobile device needs to operate. There are also a few special features
that are required to be supported by a mobile OS, but are not present in
traditional operating systems.

9.3.1  Special Constraints

The operating system for a mobile device needs to function in the presence
of many types of constraints which are not present in a traditional computer.
As an example of such a constraint, consider the fact that a mobile device
is powered by severely limited energy stored in a tiny battery. Therefore,
an important constraint for a mobile device lies in avoiding complex
computations and hence entering into a low power sleep mode as soon as
possible. Such a constraint is not usually imposed on a traditional operating
system. Further, the number of times that a mobile device is typically
turned on per day is significantly higher than that of a desktop or any
other computer. Due to this constraint, the mobile OS would need to be
loaded (booted) much faster each time after it is switched on, compared to
a desktop. Consequently, the kernel of a mobile OS needs to be of a very
small size. There are several such constraints which influence the design
of a mobile OS. It can be argued that coping with such constraints is an
important reason why the operating system of a mobile device needs to
differ significantly from a general-purpose operating system. We discuss
below some of the important constraints of a mobile OS.

Limited memory
A mobile device usually has much less permanent and volatile storage
compared to that of a contemporary desktop or laptop. To cope with the
limited memory of a mobile device, the OS must be as small as possible
and yet provide a rich set of functionalities to meet user demands. The
size of the kernel is, therefore, considered to be a very important figure of
merit of a mobile OS.
170  Fundamentals of Mobile Computing

Limited screen size

The size of a mobile handset needs to be small to make it portable.
This limits the size of the display screen. Consequently, new innovative
user interfaces need to be supported by the mobile OS to overcome this
constraint and minimize user inconveniences. For example, many handsets
provide easy configurability of the interface to suit individual preferences,
switching between menu and iconic interfaces, etc.

Miniature keyboard
Mobile handsets are either provided with a small keypad or the small-sized
display screen is designed to be used as a keyboard in a touchscreen mode
using a stylus. In both these arrangements, typing in the documents and
entering the string commands is difficult. This mandates the provision of
some facility for word completion prompts and availability of capabilities
for free form handwriting recognition.

Limited processing power

A vast majority of modern mobile devices incorporate ARM-based
processors. These processors are certainly energy efficient, powerful,
and cheaper compared to the desktop or laptop processors, yet these are
significantly slower. The sizes of the on-chip and off-chip memory are also
restricted. The cope with the restricted processing power, storage, and
battery power, usually the operating system is made to provide only a
limited number of functionalities that are useful in the actual operation of
the mobile. Activities such as mobile application development that require
use of memory-intensive utility programs, such as editors and compilers,
are carried out on a desktop or laptop, and only after the application is
completely simulated and tested, it is cross-compiled and downloaded
onto the mobile device.

Limited battery power

Mobile devices need to be as lightweight as possible to increase their
portability. Due to the severe restrictions that are placed on their size and
weight, a mobile device usually has a small battery and often recharging
cannot be done as and when required. In spite of the small battery, a mobile
phone is expected to support long talk time without the need to recharge
frequently. Consequently, the operating system for a mobile device needs
to be not only computationally efficient, but also at the same time expected
to minimize power consumption. The techniques used by an OS to reduce
power consumption include putting the processor and display screen into
sleep mode within a few seconds of inactivity, and varying the intensity
of transmitted antennae power as per requirement, etc.
 Operating Systems for Mobile Computing   171

Limited and fluctuating bandwidth of the wireless medium

The operating system of a mobile handset needs to run complex protocols
due to the inherent problems caused by mobility and the wireless medium.
A wireless medium is directly susceptible to atmospheric noise, and thereby
causes high bit error rates. Further, the bandwidth of a wireless channel
may fluctuate randomly due to atmospheric noise, movement of some
objects, or the movements of the mobile handset itself. This can show up
as short-term fades. There can be relatively longer-term disconnections
due to handoffs. In this context, uninterrupted communication requires a
special support for data caching, pre-fetching, and integration.

9.3.2  Special Service Requirements

Several facilities and services that are normally not expected to be supported
by a traditional operating system, are mandated to be supported by a
mobile OS. We identify a few important ones in the following.

Support for specific communication protocols

Mobile devices are often required to be connected to the base station
and various types of peripheral devices, computers and other mobile
devices. This requires enhanced communication support. The types of
communication protocols used for communication with the base station
depend on the generation of the communication technology (1G, 2G, etc.)
in which the mobile device is deployed. Considering that a mobile should
be usable across the existing technology spectrum, it becomes necessary
to simultaneously support two or more generations of technology. For
communication with other devices and with computers, TCP/IP and
wireless LAN protocols also need to be supported. For web browsing as
well as communication with other personal devices such as pen drive and
headphones, though mobile devices are equipped with USB and other
types of ports, mobility constraints often make infrared or Bluetooth
connections preferable. This mandates the operating system to support
multiple interfacing protocols and hardware interfaces.

Support for a variety of input mechanisms

A miniature keyboard forms the main user input mechanism for an
inexpensive mobile device. Sophisticated mobile devices (smartphones)
usually support the QUERTY keyboard. Many recent mobile devices also
support touchscreen or even stylus-based input mechanisms along with the
handwriting recognition capability. The different input mechanisms to be
supported strongly influence the intended primary use of a device as well
as the specific customer segment for which it is positioned. These issues
dictate the choice and complexity of the user interaction part of the OS to
a large extent and to a smaller extent the internal design of the operating
172  Fundamentals of Mobile Computing

system. A mobile OS needs to support a variety of input mechanisms to

make it generic and usable by different manufacturers of mobile devices.

Compliance with open standards

Adhering to an open standard facilitates the development of innovative
applications by third-party developers. To facilitate the third party software
development as well as to reduce the cost of development and time-to-
market by the mobile handset manufacturers, the OS should adhere to
open standards. Smartphones come in many different shapes and sizes and
have varying screen sizes and user input capabilities. Therefore, the user
interface and networking capabilities of a mobile OS need to be designed
keeping these diversities in view.

Extensive library support

The cost-effective development of third party applications requires extensive
library support by the OS. At the minimum, the expected library support
includes the availability of programmer callable primitives for email, SMS,
MMS, Bluetooth, multimedia, user interface primitives, and GSM/GPRS
functionalities.

9.4  A Survey of Commercial Mobile Operating Systems

It is a challenging task to design a mobile OS with a set of core capabilities

that are expected to be supported by mobile devices and with a consistent
programming environment across all smartphones that install the OS. The
mobile OS has to also facilitate third party development of application
software and yet allow manufacturers of different brands of mobile devices
to build their choice set of functionalities for the users. A few popular
mobile OS are discussed below.

9.4.1  Windows Mobile

Before discussing the important features of the mobile operating systems

available from Microsoft, we first give a historical perspective. Considering
the proliferation of embedded devices, Microsoft Corporation developed
an operating system in the year 1996 targeted specifically at these devices.
Since then, this operating system has undergone several enhancements and
modifications over successive generations. The main feature of Windows
CE operating system, which sets it apart from other traditional operating
systems, is the support that it provides for deterministic scheduling of
time-constrained tasks. Based on the Microsoft’s Windows CE operating
system, the company designed the Pocket PC 2000 operating system in the
year 2000. It was targeted for PDAs and not mobile phones.
 Operating Systems for Mobile Computing   173

Since the usage of mobile phones was increasing at a brisk rate,

Microsoft introduced its Windows Mobile operating system in the year 2003.
It was developed based on Pocket PC 2000 and was targeted specifically
as an operating system for mobile phones which the different cell phone
vendors can use in their cell phones. Unlike Apple’e iOS and RIM’s
Blackberry OS which are essentially confined to use on Apple’s iPhone and
the Blackberry, Microsoft intended the Windows mobile operating system
to be used across a wide cross section of mobile phone manufacturers. Since
the Windows Mobile operating system was developed to be suitable for use
on multiple vendor platforms, Microsoft defined a hardware specification
for hand-held computers that can run its Windows Mobile operating system
in order to simplify the design of the operating system and to reduce the
number of versions of the operating system. It was also intended to make
the cell phones manufactured by different vendors appear uniform.
Microsoft later renamed its Pocket PC operating system to Windows
Mobile Classic. Windows mobile classic operating systems support
touchscreen-based user interface but do not support any phone capability.
Consequently, Windows Mobile is now a family of three operating systems:
Windows Mobile Standard and Windows Mobile Professional are targeted for
use in smartphones, and Windows Mobile Classic is not targeted for cell
phones, but for PDAs. The evolution of Windows Mobile operating system
is schematically shown in Figure 9.2.

Windows
Phone 7
2010

Standard
Windows Professional
Mobile
Classic
2003

Pocket
PC
2000
Windows CE Windows CE
1.0 2.0
1996 1997

Figure 9.2  Evolution of Microsoft’s mobile OS.

Windows mobile is designed to have a look and feel that is very similar
to the desktop version of Windows. Microsoft obviously planned that
since many users are familiar with the desktop version of Windows, they
would find it easy to operate Windows mobile. Besides the core capabilities
required by a mobile device, many third-party software applications
are available for Windows mobile. These software applications can be
174  Fundamentals of Mobile Computing

purchased via the Windows Marketplace for mobiles. The marketplace is

a website maintained by Microsoft, where different application developers
can submit their applications for download by the subscribers. Microsoft
passes on 70% of the fee received to the developers hosting their applications.
Windows mobile has recently been superseded by Windows Phone 7.
A significant attempt to increase the market penetration of Windows
mobile was the joint announcement in 2011 by Microsoft and Nokia of
a partnership between their companies. They announced that Windows
Phone 7 operating system would be used as the operating system for Nokia
smartphones. In the press conference, they suggested that the mobile OS
marketplace would henceforth see a three horse race and named Android
and iOS, the main competitors of Windows Phone 7.
Windows Phone 7 is a significant improvement over the Windows
mobile operating system. Possibly because of this, Microsoft has not
maintained backward compatibility with the Windows mobile operating
system, meaning that a mobile application that runs on the Windows
mobile may not run on the Windows phone OS. Microsoft has defined
the hardware specifications that a Windows Phone 7 device must meet.
For example, it should support a screen resolution of 800  480 pixels.
Windows Phone 7 devices need to have an accelerometer and a compass.
Windows phone operating system provides a touchscreen interface with
facilities for both command and text input. The operating system detects
when a device has been rotated from portrait to landscape orientation.
A few important features of the Windows mobile OS are the following:
• The Graphics/Window/Event manager (GWE) component handles
all input and output.
• Provides a virtual memory management.
• Supports security through the provision of a cryptographic library.
• Application development is similar to that in the Win32 environment.
This is considered advantageous since many programmers have
knowledge of Win 32-based application development.
• At present, it does not provide true multitasking. An application in
the background goes into hibernation and gets active only when it
comes to foreground. However, it is expected that Microsoft may
support true multitasking in the future versions of the Windows
Phone operating system.

9.4.2  Palm OS

Palm OS (also known as Garnet OS) is a proprietary operating system

that was developed by Palm Computing in 1998 for its highly successful
PDA called Palm Pilot. Palm OS was designed for ease of use with the
provision of a touchscreen-based graphical user interface. Later, Palm OS
was upgraded to facilitate installation in several different mobile devices,
such as smartphones of different makes, wrist watches, hand-held gaming
 Operating Systems for Mobile Computing   175

consoles, bar code readers and GPS devices. About a decade ago, Palm
OS was a very popular operating system, but has now lost its dominant
market position (Lin and Ye, 2009).
The key features of the current Palm OS (named Garnet) are the
following:
• It is essentially a simple single-tasking operating system. As a result,
only one application can run at a time. The implications of this
are many and easily noticeable. For example, if you are on voice
communication, you cannot use the calculator, or read an SMS.
• It has an elementary memory management system. To keep the
operating system small and fast, Palm OS does not isolate the
memory areas of applications from each other. Consequently, any
misbehaving application can crash the system.
• Palm supplies Palm emulator, which emulates the Palm hardware
on a PC. This allows Palm programs to be developed and debugged
on a PC before being run on the Palm hardware.
• It supports a handwriting recognition-based system for user input.
• It supports a facility called HotSync technology for data
synchronization with desktop computers.
• It supports sound playback and recording capabilities.
• It incorporates a very simple and rudimentary security model in
which a device can be locked by password.
• The different interfaces supported include Serial port/USB, infrared,
Bluetooth and Wi-Fi connections.
• It uses a proprietary format to store calendar, address, task and note
entries and yet are accessible by third-party applications.

9.4.3  Symbian OS

Symbian operating system was developed through a collaboration among

a few prominent mobile device manufacturers including Nokia, Ericsson,
Panasonic, and Samsung. Their objective was to develop a single industry
standard operating system (Hall and Anderson, 2009). For many years,
Symbian was the undisputed leader in the smartphone OS market. It
was the operating system used in the handsets manufactured by Nokia,
Ericsson, Panasonic, and Samsung. Though Symbian had a vision of a
single collaborative effort towards the realization of a versatile mobile
OS installable on a wide range of smartphones, in reality, the different
participants tried to exert control over the direction of evolution of the
operating system and consequently it became huge and unwieldy.
In 2008, Ericsson, Sony, Panasonic, and Samsung pulled out of the
collaboration, selling their stake to Nokia. Around the same time, Google
announced Android as an open operating system. After this, the market
share of Symbian started coming down drastically. To counter this, the
Symbian source code was published under Eclipse Public License (EPL)
176  Fundamentals of Mobile Computing

in February 2010. This event was reported to be the largest codebase

transition from proprietary to Open Source in the entire history. Due to the
intensifying competition in the mobile handset market, Nokia announced in
2011 that it would move away from Symbian and use Windows Phone 7
OS in its smartphone handsets. This event is likely to cause further drastic
erosion to the Symbian installation base.
Symbian OS is a real time, multitasking, pre-emptive, 32-bit operating
system that runs on ARM-based processor designs. The inherent design of
the Symbian operating system is microkernel-based. The CPU is switched
into a low power mode, when the application is not responding to an event.
Symbian comes in two major flavours. These flavours have been developed
based on demands from various service providers.

(a) Series 60:  The series-60 platform was until recently the leading
smartphone platform in the world. The relatively large sized colour
screen, easy-to-use interface and an extensive suite of applications
make it well-suited to support advanced features such as rich content
downloading and MMS (Multimedia Messaging Service). Series
60 was mainly being used on Nokia’s smartphones and Samsung
handsets.
(b) UIQ interface:  UIQ (earlier known as User Interface Quartz) is a
software package developed by UIQ Technology for Symbian OS.
Essentially, this is a graphical user interface layer that provides
capabilities for third-party application developers to develop
applications and effortlessly create user interfaces.
A few other important features supported by the Symbian operating
system are given below:
• It supports a number of communication and networking protocols
including TCP, UDP, PPP, DNS, FTP, WAP, etc. For personal area
networking, it supports Bluetooth, InfraRed and USB connectivity.
• It supports pre-emptive multitasking  scheduling and  memory
protection. Symbian is a  microkernel-based operating system.
• CPU is switched into a low-power mode when the application is
not responding to an event.
• It is optimized for low-power and memory requirements. Applications,
and the OS itself, follow an object-oriented design paradigm.
• All Symbian programming is event-based, and the CPU is switched
into a low-power mode when the applications are not directly
dealing with an event. This is achieved through a programming
idiom called  active objects.
• Carbide is an Integrated Development Environment (IDE) toolkit
that is available for C++ application development on Symbian
OS. It essentially works as an Eclipse plug-in and contains editor,
compiler, emulator, libraries and header files required for Symbian
OS development. Development kits are available at Nokia and the
Symbian Foundation websites.
 Operating Systems for Mobile Computing   177

9.4.4  iOS

In January 2007, Apple unveiled its sleek innovative mobile device—the

iPhone—causing a storm in the smartphone marketplace. The iPhone was
designed to replace Apple’s highly successful iPod. Apple had developed
iOS as iPhone’s operating system and was originally known as iPhone OS,
but later renamed iOS. iOS is a derivative of Mac OS. Mac OS was later
extended for use in other Apple devices such as the iPod touch, iPad and
Apple TV.
iOS is a closed and proprietary operating system fully owned and
controlled by Apple and not designed to be used by various mobile phone
vendors on their systems. Apple does not license iOS for installation on
third-party hardware. However, the overwhelming popularity of iPhone
has given iOS a significant market presence. It provided several innovative
features that grabbed the market attention. For example, user interactions
with OS include gestures such as swipe, tap, pinch, and reverse pinch, all
of which have specific definitions within the context of the iOS operating
system. The other innovative user interactions that are supported by the
iOS include internal accelerometers used by some applications for shaking
the device as the undo command, rotating the device in three dimensions
to switch the display mode from portrait to landscape, etc.

9.4.5  Android

To understand the genesis of Android, we need to dig into a bit of history.

The leading search engine company Google’s income is dependent on
the average number of searches performed per day by people using its
search engine. This is because advertisers pay to a web search company
based on the average number of user hits that are recorded on its search
engine. Prior to Android, Google had no means of directing mobile users
to its search site. Even if a mobile owner wanted to search using Google’s
search engine, the prevalent operating systems of the mobile devices did
not facilitate this. For example, Verizon wireless directed customers to its
own search engine, as Verizon did not want to miss the extra revenue from
this source. Google took a serious view of this situation, since market trend
analysts were already predicting that searches from mobile handsets would
out pace those from conventional searches. Google was eagerly looking for
ways to address this problem. In 2005, Google acquired a small startup
company called Android, which was developing an operating system for
mobile devices based on Linux.
Google set up the Open Handset Alliance in 2007. It is a group of
82 technology and mobile communication companies that are collaborating
to develop the Android operating system as an open source software
for mobile devices. It facilitated any application developer to write
Android applications. Android soon became feature-rich due to its open
178  Fundamentals of Mobile Computing

source nature. Google could embed its search engine into Android, the
way Internet Explorer is embedded into Windows. By giving away the
operating system free and focusing on the search advertisement revenue,
Google successfully formed a formidable challenge to Windows mobile
and Symbian. A Gartner survey indicates that by the third quarter of 2011,
Android held 52% share of the mobile OS market compared to that for
Symbian’s 17% and iOS’s 15%.
Starting from 0% market share, in 2008, when the first mobile phone
incorporating Android was announced, it has become the dominant
player and has shown a remarkable rate of growth in market share and
user acceptance. To understand the success of Android, it is important
to understand the difficulties that users were experiencing with other
operating systems and how Android has helped overcome those.
• Since mobile manufacturers use different user interfaces and
interaction styles for their mobile devices depending on the customer
segment targeted, Android provided the ability to seamlessly use
either a phone-based keyboard or a touchscreen.
• Mobile users expect to browse real web pages, and not the simplified
mobile versions of those pages. Many mobile handsets support
browsing alternative sites provided by many website operators
for mobile handsets with small screens and limited interfacing
capabilities. However, Android scores over those operating systems
by providing a built-in full web browser capable of rendering full
web pages and not just small mobile versions.
• An important handicap of the competing operating systems is the
difficulty of development of third-party applications. For example,
an application on iOS has to be approved by Apple before it can
be offered to the users as outlined in the SDK agreement. Apple
does not, in fact, facilitate third party application development and
is implicitly promoting a closed proprietary environment, where
the internal working of the operating system is not exposed to the
developers.
• A prominent advantage that Android holds out is that Android
SDK works in Eclipse environment. Since many developers are
already exposed to these standard technologies, there is a large
pool of developers available for working on projects on the Android
platform.
• It provides an RDBMS SQLite for data storage and data sharing
across various applications.
• It has several innovative pre-installed applications such as Gmail,
Maps, voice search, etc.
Android allows application developers to write code in the Java language. It
facilitates the development of applications with the help of a set of core Java
libraries developed by Google. Though most applications are developed
using Java, applications and libraries can still be written in  C  and other
 Operating Systems for Mobile Computing   179

languages. The Android code is structured into four different layers as

shown in Fig. 9.3. The important features of these layers are discussed
below.

Figure 9.3  Android software stack.

Application layer
The Android operating system comes with a set of basic applications such
as web browser, email client, SMS program, maps, calendar, and contacts
repository management programs. All these applications are written using
the Java programming language J2ME. Android applications do not have
control over their own priorities. This design is intentional and is intended
to help aggressively manage resources to ensure device responsiveness and
even kill an application when needed.

Application framework
An application framework is used to implement a standard structure for
different applications. The application framework essentially provides a
set of services that an application programmer can make use of. The
services include managers and content providers. Content providers enable
applications to access data from other applications. A notification manager
allows an application to display custom alerts on the status bar.

Libraries and runtime

The available libraries are written using multiple languages such as C and
C++. These are called through a Java interface. These include a Surface
Manager (for compositing windows), 2D and 3D graphics, Media Codecs
like MPEG-4 and MP3, an SQL database SQLite and the web browser
engine called WebKit.
The Android runtime consists of two components. A set of libraries
provides most of the functionalities available in the core libraries of the
Java language. The other runtime is the Dalvik virtual machine. Most
applications that run on Android are written in Java. Dalvik translates
a Java application program into machine code of the mobile device and
executes it by invoking the operating system. These can be compiled
to  ARM  native code  and installed using the Android native development
kit (SDK).
180  Fundamentals of Mobile Computing

Dalvik VM is not a traditional JVM, but a custom VM designed to run

multiple instances efficiently on a single device. Every Android application
runs its own process with its own instance of the Dalvik virtual machine.

Kernel
Android kernel has been developed based on a version of  Linux kernel.
However, it has excluded the native  X Window System and does not
support the full set of standard  GNU  libraries. Obviously, this makes it
difficult to reuse the existing Linux applications or libraries on Android.
Based on the Linux kernel code, Android implements its own device
drivers, memory management, process management and networking
functionalities. Android is multitasking and allows applications to run
concurrently. For example, it is possible to hear music and read or write
an email at the same time. This layer is the one that is commonly used by
the cell phone users.
Google initially maintained the kernel code they contributed to in
the Linux public distribution. Since 2010, Google no longer maintains its
Android kernel extensions in the Linux public distribution. Now Google
maintains its own code tree. This has marked the branching of Android
from Linux code in the public distribution.

9.4.6  Blackberry Operating System

Blackberry operating system is a proprietary operating system designed

for BlackBerry smartphones produced by Research In Motion Limited
(RIM). Being a proprietary operating system, details of its architecture
have not been published. But, at the user level, the very good email
system that it deploys is easily noticed. It supports instant mailing while
maintaining a high level of security through on-device hardware-based
message encryption

9.5  A Comparative Study of Mobile OSs

A comparative summary of the important features of three popular

mobile OSs has been presented in Table 9.1 As can be seen, most of the
features provided by the three leading mobile OSs—Android, Symbian,
and Windows Phone—are more or less comparable. All these operating
systems have very small footprint, run on ARM-based processors, and
support demand paging. What needs to be borne in mind is that a vast
majority of handsets are designed for ARM-based processors.
There are, however, many subtler points on which they differ. For
example, both iOS and Windows Phone 7 are proprietary operating systems
and not many details about the internal design of these two operating
 Operating Systems for Mobile Computing   181

systems are available. Even iOS does not facilitate the development of
third-party applications. However, this has a downside. The design and
development of a good operating system by a vendor is not sufficient to
warrant its wide customer acceptance. It is advantageous to have a variety
of applications available to run on an operating system. As has happened
earlier in the computing history, it becomes a losing game for an operating
system vendor, if it relies on itself alone to develop a large number of
innovative applications on its operating system. In this regard, Android
and Symbian being open operating systems, score over the other market
players.

TABLE 9.1  A Comparison of the Features of Three Popular Mobile Operating

Systems

Feature Android Symbian OS Windows

Phone 7
License Public, Free, and Initially was private, Proprietary
Open Source later became public.
Footprint 250 KB 200 KB 300 KB
Change of UI Possible No No
Power management Yes Yes Yes
Kernel Linux with minor Proprietary Win CE
changes
True multitasking Yes Yes No
Premeptive Yes Yes Yes
scheduling
Demand paging Yes Yes Yes
CPU architecture ARM, MIPS,  86 ARM ARM
supported

Android provides a flexible UI and is rich in features, being based on

the open source Linux. Windows Phone lacks many features provided by
Android, and iOS lacks many features even when compared to Windows
Phone. For example, it lacks features such as Bluetooth file transfers, file
manager, widgets and FM radio, etc.
Besides the above mentioned points, another sore point with the
Windows Phone operating system is that it expects the mobile device
hardware manufacturers to make their devices compliant to the Windows
Phone specification to make them uniform. This is something that the
manufacturers of mobile devices do not like as it makes it difficult for
them to provide innovative features to attract new customers.
A major reason behind Android’s success is that it facilitated
competitiveness of hardware makers without good software capabilities.
182  Fundamentals of Mobile Computing

The royalty-free Linux-based Android has been adopted by all the major
Asian handset makers such as Samsung, HTC, LG, etc. as well as Motorola
and Sony Ericsson (Farooq and Kunz, 2011).

9.6  Operating Systems for Sensor Networks

A sensor node is tiny and needs to operate in an extremely power-

constrained environment. Consequently, it deploys a rudimentary operating
system and does not have a kernel mode of operation. It does not support
dynamic memory allocation nor does it support virtual memory. It also
does not use tasks, signals, and exceptions, but uses functional call in its
place. A schematic of the structure of a sensor operating system is shown
in Fig. 9.4. Observe that the scheduler invokes the different application
components in response to a specific event.

Scheduler Sensor OS

Application Application Application

Component Component Component

I/O COMM. Clock

Figure 9.4  Schematic model of the structure of a sensor operating system.

The important operating systems that are available for sensor nodes
include TinyOS, Contiki OS, Lite oS, and MANTIS (Farooq and Kunz, 2011).

SUMMARY

An operating system is a vital component of any modern mobile handset,

especially the high-end ones. The increasing sophistication of mobile OS
has made it much easier and less expensive to develop useful mobile
applications and integrate them into the phone system. The availability
of numerous innovative applications in turn fuels the demand for the
smartphones. In this chapter, we first identified the special features
required of a mobile OS and the constraints under which these work. We
then surveyed and examined the important features of popular commercial
mobile OSs.
The marketplace for mobile OSs is at present fragmented and the
competition among the various popular OSs is intense. Considering that the
features provided by mobile operating systems and the market dynamics
are both evolving at a rapid pace, it can be argued that the market shares
 Operating Systems for Mobile Computing   183

of various operating systems is also rapidly fluctuating based on the

novel features they are introducing and at the same time a host of several
other commercial factors are also coming into play. About a decade ago,
Microsoft’s Phone operating system was the leader. Later, the leadership
position was wrested by Symbian. Now, Android appears to be doing the
same. Amid the chaos in the operating system market, even the near future
is hard to predict.
At present, each operating system vendor is focusing on offering new
and innovative features and no standardization has emerged. This is in
contrast to the desktop and server operating systems, where POSIX is an
accepted standard. In the absence of standardization, not only the users
find it difficult to use phones from different manufacturers, but even
the application developers and the smartphone hardware makers face
considerable challenges.

FURTHER READINGS

Farooq, Omer Muhammad and Thomas Kunz, “Operating Systems for

Wireless Sensor Networks: A Survey,” Sensors, www.mdpi.com/
journal/sensors, 2011.
Hall P., Sharon and Eric Anderson, “Operating Systems for Mobile
Computing,” Journal of Computing Sciences, Vol. 25, Issue 2, pp. 64–71,
December 2009.
Kenney, Martin and Bryan Pon, “Structuring the Smartphone Industry:
Is the Mobile Internet OS Platform the Key?,” Journal of Industry,
Competition, and Trade, Springer, Vol. 11, pp. 239–261, 2011.
Lin, F. and W. Ye, “Operating Systems Battle in the Ecosystem of
Smartphone Industry,” IEEE International Symposium on Information
Engineering and Electronic Commerce, pp. 617–621, 2009.
Mall, R., Real Time Systems: Theory and Practice, Pearson Education, 2008.
Vaughan-Nichols, S.J., “OSs battle in the smartphone market,” IEEE
Computer, Vol. 36, No. 6, pp. 10–12, June 2003.

EXERCISES
1. Explain the special features that an operating system for a mobile
device needs to support compared to the features provided by a
traditional operating system.
2. What is a microkernel operating system? Why is microkernel-based
design being preferred for developing a mobile OS?
184  Fundamentals of Mobile Computing

3. How is the operating system for a mobile phone any different

from the operating system for a desktop? Name a few important
commercial mobile operating systems.
4. Name three commercial operating systems for mobile phones.
Outline the important functionalities supported by any one of them
which are not supported by a traditional operating system.
5. Explain the principal functions of the operating system of a mobile
device. Discuss how an example application can be implemented
on a mobile device and the specific operating system services that
it makes use of.
6. Compare the features provided by the following mobile operating
systems: Android, Symbian, and Windows Phone 7.
7. Discuss the architecture of the Android operating system. Briefly
identify the possible reasons as to why it has been able to rapidly
improve its market share compared to its peers since its introduction
a few years ago.
8. Discuss the important features provided by a modern mobile
operating system, the special constraints under which a mobile
OS needs to function, and the future trends in the development of
mobile operating systems.
9. Using at least one suitable example, explain the flexibilities that a
user would be required to sacrifice when a single-tasking operating
system is used in the mobile device.
10. What is a microkernel operating system? What are its advantages
and disadvantages in the context of the design of a mobile operating
system?
11. Briefly explain how an operating system for a sensor network is
different from a traditional operating system.
 Mobile Application Development
10
and Protocols

W ith the rapid increase in the popularity of mobile computing

applications, several tools, techniques, and standards and application
layer protocols have become available for developing these applications.
As already mentioned, many mobile computing applications are essentially
web-based client-server programs. However, mobile application
development is considerably different from the traditional client-server
program development. The limitations inherent to the mobile wireless
networks and mobile handsets pose several challenges to the development
of mobile computing applications using the traditional client-server
programming techniques. In this chapter, we discuss various protocols
and application development frameworks that have been proposed for
mobile computing application development.
We first discuss the specific deficiencies of the mobile devices
and networks that surface when web-based client-server application
developments are attempted on these. Subsequently, we outline the broad
approaches that have been proposed to circumvent these deficiencies.
Next, we discuss the present technologies of WAP and Bluetooth. Finally,
we discuss J2ME and Android-based mobile application development
frameworks.

10.1  Mobile Devices as Web Clients

The World Wide Web (WWW) was originally designed to work with
traditional computers such as desktops and laptops. The WWW makes
implicit assumptions regarding the capabilities of the web clients (desktops
and laptops) and the network over which these are run. Considering that
the early mobile networks supported only voice traffic and the mobile
handsets had rather primitive computing infrastructures and had screens
185
186  Fundamentals of Mobile Computing

with only text display capabilities, it was preposterous to even think

of web access from mobile handsets. As mobile handsets and mobile
networks gradually evolved over time, the mobile networks could support
data traffic and the handsets became reasonably powerful. This prompted
practitioners and researchers to explore the feasibility of invoking web-
based client-server applications from mobile handsets. Even the present-
day high-end mobile devices fall far short of the capabilities of the
traditional computers on many counts and the latest wireless mobile data
networks such as 3G provide data rates that are many times lower than that
available to traditional web clients. Another dimension of difficulty arises
on account of the frequent short duration signal fading and disconnections
that mobile clients (we shall refer to mobile handsets running web-based
mobile applications as mobile clients) are susceptible to. It had become
clear to the researchers that unless these inherent deficiencies of mobile
devices and networks were properly addressed through the development
of appropriate protocols and techniques, web access using mobile devices
would be frustrating and meaningless.
As we have just pointed out, three main problems need to be addressed
before a mobile user can meaningfully access the web and run web
applications using a hand-held device. One of the problems pertains to
the extremely slow network speeds that mobile devices have to do with—at
present network access speeds are restricted to only a few kilobytes per
second. In contrast, the typical bandwidth available for web surfing using
traditional desktops is a few gigabytes per second or at least several mega
bytes per second. This problem of low bandwidth of mobile networks can
show up as excruciatingly slow web access to the mobile users. Another
difficulty of using mobile devices as web clients, is the small screen that
these devices have. A typical screen of a mobile device would not be able
to satisfactorily display even a reasonable portion of a normal web page. It
is simply not possible to squeeze a desktop-sized screen into the palm-sized
screen of a mobile device. Further, the computing capabilities and memory
sizes of the hand-held devices are severely limited. If the conventional web
pages are to be directly downloaded and displayed, without incorporating
any remedial measures to compensate for the miniature screens, the user
would find them unusable and get utterly frustrated. The third problem
concerns the disconnections and signal fading that occur when a user
moves around, rendering any meaningful usage of the Internet futile.
In order to address the above three problems, a large number of
approaches have been suggested in recent years. The essential idea behind
all the approaches proposed to help provide meaningful web access from
mobile hand-held devices is simple. The two problems of low bandwidth
and small screen size can be made to solve each other. If a device has a
small screen and can only display text and simple graphics, high-speed
connectivity is not needed either. With this overall solution scheme in
mind, let us now investigate the genesis of the problem. The founders of the
WWW (Tim Burners-Lee and other scientists) had envisaged that HTML
 Mobile Application Development and Protocols   187

pages can be seamlessly displayed across a multitude of devices of greatly

varying capabilities. In the pure form of this objective, this can enable
people to satisfactorily access the web using their respective mobile devices.
However, in the later years, the development of dynamic web content
and other advancements made the web move away from this ideal and
slowly led to HTML pages incorporating sophisticated graphics, animation,
and multimedia contents that implicitly assumed the availability of high
bandwidth network access and the presence of significant processing power
at the web clients.
After the WWW had an unprecedented success and almost unbelievable
user acceptance, the World Wide Web Consortium (W3C) was set up
by Tim Berners-Lee (the founder of WWW) at MIT. W3C is a nonprofit
group that oversees the formation of various Web standards. Considering
the vast popularity of mobile phones and the imminent possibility of
web-based information access and application execution from mobile
devices, it announced in 1998 the creation of a special version of HTML
for mobile devices. This special version of HTML was called compact
HTML (C-HTML). In this version of HTML, the advanced features of web
such as fonts, frames, tables, graphics, and dynamic content were omitted
with the intent of not only saving bandwidth but also freeing the hand-
held devices of computational overload. However, care was taken to allow
the use of several essential and standard web technologies such as SSL
(Secure Socket Layer) for secure communication. Needless to say that this
is a rather simple implementation of the overall solution schema that we
have pointed out and does not take into account many issues that plague
satisfactory web access by mobile clients. C-HTML underwent substantial
overhaul and improvements over the next few years and helped define the
present web access technology in mobile computing environments. In this
chapter, we briefly discuss the progressive evolution of technology that has
now made it possible for mobile clients to meaningfully access the web.

10.1.1  HDML (HanDheld Markup Language)

HDML was developed by a company called Unwired Planet. Unwired

Planet demonstrated that by using a micro-browser developed by it, web
pages written in HDML can be meaningfully accessed from a mobile
handset. Like C-HTML, HDML required that a web page should avoid the
use of complicated features such as tables, frames, dynamic content, etc.
Hence it replaced a web page with two text layout metaphors: cards and
decks. A card is defined as a single user interaction, which in HTML terms
would be a web page. However, the analogy of a card with a web page is
not exact. For example, a card can contain either one specific information
display, a data entry form, or a choice menu. A web page can have many
combinations of these basic types of web pages. A single HDML file
consists of many cards. So, it is called a deck. An advantage of the card
188  Fundamentals of Mobile Computing

and deck model in the context of mobile networks is that a click on a web
page causes downloading of the entire deck associated with the web page
on to the mobile device. This substantially reduces the latency of access.
HDML eventually formed the basis of the WAP standard.
For facilitating the development and portability of applications over
wireless communication networks, telecommunication companies, Ericsson,
Nokia, Motorola, and Unwired Planet founded the WAP forum. The
mandate for WAP forum was to develop technologies that would become
the de facto standards and would help in the formulation of standards by
the appropriate standards bodies. Later the WAP forum became a part of
the Open Mobile Alliance (OMA) which has a much wider membership
and mandate. OMA is a standards body which develops open standards
for the mobile phone industry. In 1998, WML (Wireless Markup Language)
was announced. Both HDML and WML share the same basic programming
model and functionality. However, the main difference between them is
that while WML is XML-based, HDML is not. An important benefit of
WML being XML-based is that WML is a widely accepted standard for
data interchange. Further, a company can use commercially available XML
tools to generate, parse and manipulate WML, and they can also use XSL/
XSLT1 to construct WML decks from XML meta-languages.

10.2  WAP

WAP (Wireless Application Protocol) has now become the de facto

standard technology for web access by a mobile client. WAP helps wireless
applications to seamlessly access the Internet. WAP differs from C-HTML
and HDML in that it is not a markup language—it is a complete stack of
protocols. It has been designed not only to overcome specific problems
caused by wireless networks during web access (such as high latency and
jitter) but also to address those that are caused by the mobile devices such
as small screen size and low processing power. This makes WAP much
more complicated than HDML or C-HTML, but make it more effective.
In fact, the WAP architecture defines an optimized protocol stack for
communication over wireless media, a content description language, and a
miniature browser. It replaces HTML pages with highly condensed WML
pages, called cards. WML has been defined considering the limited display
characteristics of the mobile devices.
A traditional web access mechanism has been schematically shown in
Figure 10.1. As shown in this figure, the client and the application server
communicate using the http request and response streams. We now compare
this mode of web access from a traditional computer with a WAP-based

1.  XSL stands for Extensible Stylesheet Language. It is a family of languages

that can help to appropriately display XML documents. XSLT stands for XSL
Transformations.
 Mobile Application Development and Protocols   189

web access. The working of a WAP-based web access has been shown in
Fig. 10.2. In this, all requests from the mobile device are sent to the WAP
gateway. The WAP gateway transforms them into HTTP request-response
streams, and sends them to the application/content server. The reverse of
this sequence takes place when the application/content server responds to
a specific request. It is the responsibility of the WAP gateway to convert
a WAP request/response to an HTTP request/response. This conversion
is very compute-intensive. To handle this computational workload, the
WAP gateway needs to be a powerful computer. It is not difficult to
see why WAP is often described as a network-centric protocol—most
of the intelligence and computations associated with the WAP protocols
are embedded in the network rather than the simple phone. The WAP
forum released its first specification of WAP (referred to as WAP 1.0) in
the year 1997. The WAP 2.0 was subsequently released. It had backward
compatibility with the previous WAP protocol.

Figure 10.1  Traditional web access.

Figure 10.2  WAP-based web access.

Although WAP has its own protocol stack, it is designed to be

compatible with the Internet. Pages in WAP are converted to the http
and TCP protocol at the gateway. WAP 2.0 provides support for the
protocols that are counterparts of IP, TCP, and HTTP. It is also flexible
and bearer independent—meaning that WAP services can run over any
specific wireless data bearer technologies such as SMS, GSM, GPRS, etc.
The first web services ran over GSM, requiring the user to dial a specific
phone number just as one would call an ISP from a traditional phone.
190  Fundamentals of Mobile Computing

The architecture of WAP 2.0 protocol stack has been shown in

Fig. 10.3. As can be seen from the figure, the WAP protocol operates at
several layers. WAP defines its own specific protocols corresponding to
the various layers of operation of the Internet. The protocols in the WAP
stack and the bearer services in WAP are discussed below.

Figure 10.3  WAP protocol stack.

Wireless Application Environment (WAE)

WAE includes the micro-browser on the device, WML (the Wireless Markup
Language), WMLS (a client-side scripting language), telephony service, and
a set of formats for the commonly used data such as images, phone books,
and calendars.

Wireless Session Protocol (WSP)

WSP helps establish a web browsing session from a mobile handset.
WSP is based on the HTTP protocol and provides the basic session state
management, and facilities reliable and unreliable data push (called WAP
push).

Wireless Transaction Protocol (WTP)

The WTP layer in the WAP stack can be considered to be the equivalent of
the TCP layer of the TCP/IP stack, but it takes into account the availability
of low bandwidth by providing different classes of transaction services.
WTP transaction services include reliable request and response that have
been adapted to the wireless world. WTP handles the problem of packet
loss more effectively than TCP. Packet loss is a fairly common phenomenon
in wireless technologies due to factors such as atmospheric noise, signal
fading, and handoff. The packet losses are often misinterpreted by TCP as
network congestion, thereby drastically reducing the network throughput.
The problem of packet loss is effectively handled by WTP.
 Mobile Application Development and Protocols   191

Wireless Transport Layer Security (WTLS)

WTLS is the security layer that is used to transfer data securely between
a mobile device and a server. It provides support for data security and
privacy, authentication, as well as protection against denial-of-service
attacks.

Wireless Datagram Protocol (WDP)

WDP is the bottom-most protocol in the WAP protocol suite. It functions as
an adaptation layer in a wireless communication environment that makes
every data network look like UDP to the upper layers by providing services
for transport of data in the unreliable wireless environment. WDP invokes
services of one or more data bearers such as SMS, GPRS, CDMA, UMTS, etc.

Bearer Interfaces
A bearer is a low-level transport mechanism for network messages.
Considering the diversity of transport technologies, WAP is designed to
operate with SMS (Short Message Service) to GPRS (General Packet Radio
System), UMTS and IP. GPRS is relatively faster, and provides “always-
on” connections for wireless devices. WAP supports circuit-switched bearer
services such as dial-up networking using IP and Point-to-Point Protocol
(PPP). However, packet-switched bearer services are much better suited
than circuit-switched bearer services for mobile devices as they can provide
more reliable services in the unreliable wireless connection environment.

10.3  J2ME

Sun Microsystems Ltd. created the Java language in 1995. Even though
Internet had already made its appearance by 1995, but none of the available
programming languages could effectively work with Internet. Java for the
first time made it possible for programs to be run by making it possible
for web clients to download and run applets. Besides, it had several other
advantages. As we all know, Java was a resounding success and quickly
found widespread acceptance. This was followed by Java Enterprise Edition
(J2EE) for server side programming. Later Java2 Micro Edition (J2ME)
was developed for supporting programming of mobile devices. J2ME has
been targeted for use in very small devices (called J2ME devices) such as
smartphones, interactive television set-top boxes, pagers, PDAs and other
wireless devices. It includes different virtual machines and application
programming interfaces (APIs).
One of the identified design objectives for J2ME was to provide
portability across several mobile platforms, so that the same application
could be run across the mobile devices manufactured by different vendors
192  Fundamentals of Mobile Computing

and running different operating systems. Mobile hand-held devices have

very restricted capabilities compared to the traditional computers. J2ME
was designed keeping all the above constraints in mind. Some of the major
differences between conventional computers and the J2ME devices are the
following:
• Limited processing power
• Limited system memory
• Limited storage capacity
• Small display
• Low battery power
• Limited connectivity to internet
J2ME includes a miniature version of JVM called KVM (K Virtual
Machine) which can run small Java programs (called midlets) on the mobile
devices. This brought the “write once, run everywhere”, principle espoused
by Java to the mobile handsets.
The J2ME applications are lightweight and support many useful
features. A few important capabilities of J2ME programs are the following:
• Opening UDP connections between two devices.
• Establishing HTTP connections with a server
• Making Socket connections.
• Bluetooth programming .
• Bar code scanning.
Some of the popular applications of J2ME are the following:
• Automotive systems
• Set-top boxes and interactive televisions
• Network-connected consumer devices that use graphic user interface
(GUI)

10.3.1  J2ME Configuration

J2ME takes into account the fact that there are a wide variety of mobile
devices with varying capabilities and hence the need for supporting a
diverse range of features and functions. A J2ME configuration targets
to make it applicable to devices with a specific range of capabilities. A
profile on the other hand, selects a configuration and a set of APIs to
target a specific domain of applications. The selection of an appropriate
configuration and profile are therefore the crucial parts of any application
development. Since lightweight appliances do not need the capabilities of
the entire Java2 platform, by selecting the appropriate configuration and
profile, the resource requirements and the cost of running the applications
can be reduced to a large extent. The J2ME configuration parameters
include the following:
 Mobile Application Development and Protocols   193

• Availability of memory space and memory type

• Specification of processor in terms of speed and type
• Network connectivity of the device
J2ME currently defines two configurations:

Connected Limited Device Configuration (CLDC) for handheld devices

CLDC is targeted to the lower-end range of consumer electronic devices
that are designed using 16-bit or 32-bit small computing devices with
limited memory. These devices usually have between 160 KB and 512
KB of available memory. Usually powered by battery, they have low
bandwidth network wireless connections. These devices include pagers,
personal digital assistants, cell phones, dedicated terminals, and hand-
held devices. These devices use a stripped-down version of the JVM. Java
enables the purchase and download of small Java applications (called
midlets) to the handsets.

Connected Device Configuration (CDC) for plug-in devices

CDC devices are rather the higher-end devices that use a 32-bit processor,
have at least 2 MB of memory available, and implement a complete
functional JVM. CDC devices include digital set-top boxes, home
appliances, navigation systems, point-of-sale terminals, and smartphones.
The programming environment for application software development
consists of Java Virtual Machine (JVM) and a core collection of java classes.
The J2ME virtual machine could not support all the Java language features
or instructions byte codes and the software optimization provided by
J2SE (Java 2 Standard Edition) virtual machine due to the limitations in
processor and memory of low-end devices. Therefore, J2ME VMs usually
incorporate some part of the Java Virtual Machine and a subset of the Java
Language syntax.
J2ME is modular and scalable. J2ME consists of four layers of
functionalities as shown in Fig. 10.4. We explain the functionality of these
layers in the following:

Java Virtual Machine (JVM) layer:  Java Virtual Machine (JVM) is

implemented in this layer which is customized based on the capability of
a particular device and a particular J2ME configuration.

Figure 10.4  J2ME architecture.

194  Fundamentals of Mobile Computing

Configuration layer:  The features of Java Virtual Machine (JVM) and the
available Java class libraries for a specific category of devices are defined
in this layer. For profile implementers, this layer plays a vital role, but is
less visible to the users.
Profile layer:  The set of application programming interfaces (APIs) available
on a particular family of devices is defined in this layer. Applications are
written for a particular profile, and applications are portable to any device
that supports the specific profile for which it was developed. A device can
support multiple profiles. This is the layer that is most visible to users and
application providers.

Mobile Information Device Profile (MIDP) layer:  The Java APIs which
are present in this layer provide reusable functionalities corresponding to
user interface, persistence storage, and networking.

10.4  Android Software Development Kit (SDK)

Cell phones are small in size and therefore can easily be carried everywhere
like a wallet. Considering their portability and the powerful feature sets
that they provide, they have now come to play an important part in today’s
society. Now with every passing day, mobile phones implement new and
innovative functionalities and have come to possess a surprisingly large
number of functionalities. For example, many handsets provide facilities for
radio and television reception, timer and clock, Internet access, camera, and
calculator, just to mention a few. These applications are well known and
standard. These are usually developed by the operating system developers
themselves and are bundled with the operating system. However, the true
power and distinguishing features of a handset come into picture with the
third-party developed applications. Therefore, a successful mobile operating
system needs to facilitate the development of third-party applications. In this
regard, open operating systems stand out in facilitating the development of
third-party applications and Android application development tools have
now been well received.

10.4.1  Android SDK Environment

The Android SDK (Software Development Kit) is a mobile application

development framework using which developers can create applications
for the Android platform. The Android SDK provides the tools and
libraries necessary to develop applications that can run on Android-based
devices. An important advantage of Android SDK is the low processor
and RAM requirements. Besides, Android SDK can be installed on almost
all common operating systems such as Windows, Mac OS, and Linux. The
 Mobile Application Development and Protocols   195

SDK comes with an Integrated Development Environment (IDE) and other

tools which are required to develop applications. Android SDK converts
Java byte code to Android’s Dalvik VM bytecode. It is important to realize
that while developing android-based applications, the developer codes
the applications using Java but the mobile device running Android does
not finally run the Java byte code but runs the Dalvik VM byte code. For
this reason, applications developed using J2ME cannot be run directly on
Android mobile phones although they both use execution environments
that are derivations of JVM.
The environment to develop applications for Android consists of the
Android SDK, the IDE Eclipse and the Java Development Kit (JDK). After
installing the SDK, which is done by simply extracting the downloaded ZIP
file in a folder, the path to the SDK has to be set in the path environment
variable. Eclipse can be used as the IDE, which also automatically installs
the Android SDK as a plug-in. After installing this plug-in, one can start
the development of Android applications.

10.4.2  Features of SDK

Using the SDK, one can either run the application on the actual Android
device or a software emulator on the host machine. This is achieved by
using the Android Debug Bridge (ADB) available with the SDK. ADB is a
client-server program and includes three main components:
• A client program which runs on the developer’s (called host)
machine. One can invoke a client from a shell by issuing an adb
command.
• A daemon program which runs as a background process on each
emulator or device instance. It is the part that actually manages
the communication with the handset or the emulator and helps in
executing the application.
• A server program which runs as a background process on the host
machine. The server manages communication between the client and
an adb daemon that runs on the emulator or the Android handset.

10.4.3  Android Application Components

Application components are the essential building blocks of an Android

application. The following are the four components of an Android
application.

Activity:  Each activity presents a GUI screen of an application. For

example, a chat application might have one activity that allows to create
a chat, another to view the previous chat sessions, etc. Different activities
form a cohesive chat application.
196  Fundamentals of Mobile Computing

BOX 10.1  Android application development

Android originated from a small software company acquired by Google and
is now owned by the OHA (Open Handset Alliance) of which Google is a
member. The Open Handset Alliance (OHA) is a consortium of a large number
of companies. OHA focuses on the development of open standards for mobile
devices. Besides Google, the present members of OHA include HTC, Sony,
Dell, Intel, Motorola, Qualcomm, Texas Instruments, Samsung Electronics,
LG Electronics, T-Mobile, Nvidia, and Wind River Systems. Android SDK is a free
tool in the sense that anyone can use Android SDK to develop applications
to run on Android OS. Android makes use of only non-proprietary techniques
such as Java programming and extensible mark up languages. This makes it
an open tool. Google’s Software Development Kit supports a relatively large
subset of the Java Standard Edition 5.0 library. This includes the following APIs:
java.io: Contains classes for system input and output through data streams.
java.net: Contains classes for implementing networking applications.
java.security: Contains classes for the security framework.
java.util: Contains several utility classes such as event model, date/time facilities
and internationalization.
javax.sound.midi: Contains a collection of classes for input and output,
sequencing and Musical Instrument Digital Interface.
A few of the available important Android APIs are the following:
android.graphics: Contains classes for low level graphics.
android.media: Contains classes for media player and recorder functionalities.
android.net: Contains classes for network access that are not included in the
normal java.net APIs.
android.opengl: Contains classes for utilities (OpenGL) for 2D/3D graphics.
android.telephony: Contains classes that help receive and monitor phone calls.
android.database.sqlite: Contains classes that help manage a private database.
android.view: Contains classes to handle screen layout.
android.widget: Contains classes for user interface elements to use on the
application screen.
android.content: Contains classes for accessing and publishing data on the device.
android.app: Contains classes for high level classes encapsulating the overall
Android application model.

Content providers:  Content providers are used for reading and writing
data that are either private to an application or shared across applications.
By using the content provider, an application can query or modify the
stored data.
Service:  A service denotes a background task and not for interacting
through a user interface. For example, a service might play music in the
background while the user is interacting with a different application.
Broadcast receivers:  The broadcast receiver responds to broadcast
announcements by an application. For example, a battery monitoring
application might broadcast that the battery is low. Based on this, the music
player might reduce the volume or the screen display may be dimmed.
 Mobile Application Development and Protocols   197

10.4.4  Android Software Stack Structure

To a user of a mobile handset, various functionalities are provided by a

cooperative working of a number of application programs and system
programs. These collection of programs can be decomposed into a hierarchy
of four layers as already discussed in Section 9.4.5 (see Fig. 9.3). The
functionalities of these layers are also discussed in that section, which may
be referred to for ease of recollection.

10.4.5  Advantages of Android

Application development in the Android platform is becoming popular

due to the many advantages that it offers.
• The mobile platform Android is an open platform and can be ported
on almost every type of cell phone.
• The Android SDK to develop applications is possible on every
operating system.
• Android requires a low footprint of 250 KB.
• The emulator of the Android platform has a modern design and is
easy to use.
• Application installation on the emulator/device is possible via
Android Debug Bridge (adb) or via Eclipse (with ADT plug-in)
• Google offers a very good documentation as well as many examples
which cover the most basic and important techniques used to get in
touch with Android and the application development on it.
• Android supports robust libraries for media access, communication
and data transfer.
• Android offers a real database SQLite using which meaningful data
manipulation and data sharing across applications is possible.
• Android has an integrated web browser which gives an experience
similar to web browsing using a desktop PC.
• Android uses the standardized and open programming language
Java.

SUMMARY

In this chapter, we discussed the application layer protocol WAP. We also

discussed J2ME and Android SDK that facilitate application development
for the mobile phones. These come with a programming environment
installable on a Windows or Desktop PC where the application can be
developed. After the application is run on a mobile device emulator, it is
downloaded on the actual mobile device.
198  Fundamentals of Mobile Computing

EXERCISES
1. What problems would occur if an attempt is made to enable the
mobile user to browse the web by interfacing the mobile networks
and the mobile devices directly to the WWW with only trivial
extensions to the existing web protocols?
2. Why is HTML-based standard web access not very meaningful in
mobile environments? How does WAP address this issue?
3. What is C-HTML? How did C-HTML address problems that arose
while accessing traditional web using mobile handsets? What are
the shortcomings of C-HTML?
4. What is HDML? What were the problems of traditional web access
that HDML tried to solve? How have these problems been solved?
5. Explain WAP 2.0 protocol and its architecture. Briefly explain why
WAP is called a network-centric protocol.
6. Distinguish between traditional computing devices and the J2ME
devices.
7. What is J2ME? What facilities does J2ME offer to a mobile application
developer? Briefly explain the architecture of J2ME.
8. Write a short note on Android software stack.
9. Using a suitable schematic diagram, explain how a traditional web
access differs from that done using WAP.
10. Briefly write how an application can be developed using the Android
SDK.
 11
Mobile Commerce

S o far we have discussed the various layers of protocols used in mobile

computing. In this chapter, we provide an overview of mobile commerce
as an example application of the mobile computing infrastructure. Mobile
commerce (M-commerce in short) is an important application of mobile
computing. Mobile commerce, in simple words, involves carrying out
any activity related to buying and selling of commodities, services, or
information using the mobile hand-held devices. M-commerce has over
the last decade become extremely popular. The popularity of m-commerce
can be traced to the convenience it offers both to the buyers and sellers.
Though buying and selling may appear to be an obvious application
of a mobile hand-held device, an important issue in M-commerce is how
payments can be made securely and rapidly as soon as a buyer decides
to make a purchase. So far, the use of computers and networking in
trade related transactions has been limited to automatic teller machines
(ATMs), banking networks, debit and credit card systems, electronic money
and electronic bill payment systems (E-payment). Each of these modes
of transaction has its own advantages and disadvantages and warrants a
detailed investigation for use in M-commerce. Mobile payment is a natural
evolution of E-payment schemes and has found an important place in
M-commerce.

11.1  Applications of M-Commerce

M-commerce applications can be broadly categorized into either B2C or

B2B. These two categories of applications are discussed next.

199
200  Fundamentals of Mobile Computing

BOX 11.1  Evolution of M-commerce

Money is now an important element of all business and trade. In older times,
money did not exist. What existed was a simple “barter system” where things
could be exchanged, say, fish for grains, The evolution of currency (money)
gave birth to the concept of a “marketplace”. In a marketplace, commerce
is a function of 4 Ps—Product, Price, Place and Promotions. All these four
components play a vital role for a business transaction to take place. Different
combinations of 4Ps determine the different forms of commerce. Once the
marketplace came into existence, a few pioneers realized that people would be
ready to pay extra if products could be delivered at the customer’s doorsteps.
A small change to two of the Ps, Price and Place, led to the convenience of
getting products at customers’ homes. This concept delighted the customers
and thus, the concept of “Street Vendors” was born. When the postal system
came into being, sellers found a new avenue and started using mails to describe
their products. It ultimately led to the concept of “Mail Order Cataloguing”. A
mail order cataloguer buys goods and then sells those goods to the prospective
customers. A mail order catalogue is a list of the goods that the cataloguer
deals with. From this point, the evolution of the “Teleshopping” networks
was inevitable with the development of the Internet. The latest generation
of commerce is being done over the Internet. The Internet provides a virtual
platform where sellers and buyers can come in contact for sale and purchase of
goods and services, even though they may be thousands of miles apart, belong
to different countries, and might speak different languages. “E-Commerce”
has emerged as a boundary-less trade medium accelerating the pace of
globalization.
   The Internet has already reached the home of most customers. In this
context, the distribution channel has started to assume a new meaning to the
e-marketer. With options of paying online through debit and credit cards, on-
line transactions have become purely electronic. In this context, the difference
between E-commerce and M-commerce is that E-commerce is limited to PC
users with an Internet connection, while M-commerce has been adopted by
the mobile phone users.

11.1.1  Business-to-Consumer (B2C) Applications

Business-to-consumer (B2C) is a form of commerce in which products or

services are sold by a business firm to a consumer. B2C is an important
category of mobile commerce applications and is reported to be nearly half
of the total M-commerce market (Varshney et al., 2000). A few examples
of B2C applications are given below:

Advertising
Using the demographic information collected by the wireless service
providers and based on the current location of a user, a good targeted
advertising can be done. The wireless service provider may also keep
track of the history of the purchases made by customers by directing
 Mobile Commerce   201

advertisements to mobile phones. Customers may also solicit specific

advertisements. For example, suppose a consumer in a shop is fascinated
by a new electronic product and wishes to buy it but only after getting
more details about it. For this purpose, he can view all the relevant
advertisements for the product by taking the picture of the bar code using
his mobile device.

Comparison shopping
Consumers can use their mobile phones to get a comparative pricing
analysis of a product at different stores and also the prices of the related
products. For example, suppose consumers visiting a shop can use their
mobile phones to access a web-based comparison shopping application.
By scanning the bar code on a product, the consumer can see the price of
this product at different shops in the adjacent area. After seeing how the
product is priced at different shops, the consumer may decide to buy from
a shop where it is competitively priced. In a similar manner, consumers
can also access product reviews from consumer organizations or customers.

Information about a product

Consumers can access additional information about products through
their mobile phones. Assume that a consumer buys some medicine in
a pharmacy shop, but cannot read the dosage instructions on the carton
given in German and Spanish languages only. The consumer can, however,
scan the bar code on the pack using the mobile device to read the dosage
instructions in the English language, which he knows.

Mobile ticketing
Mobile phones can be used to purchase movie tickets (called m-tickets)
using credit cards. After the payment is received, a unique bar code is sent
to the purchaser’s mobile phone by an SMS. The purchase can gain entry
to the movie hall by showing the bar code downloaded into the mobile
device to a bar code reader at the entrance. Consider another example. A
customer books a train ticket using the mobile phone. An m-ticket or a
text message is sent to the mobile phone. Inside the train, by showing the
m-ticket on his mobile phone to the ticket collector, the traveller occupies
the seat.

Loyalty and payment services

In this application, mobile phones can replace the physical loyalty cards.
Having signed up for a supermarket loyalty scheme, a unique bar
code is sent to a consumer’s mobile phone. After shopping at the same
supermarket, the consumer shows the bar code at the cash counter and
accumulates points based on the total amount spent.
202  Fundamentals of Mobile Computing

Mobile phones can be used to make payments. For example, consumers

can buy canned drinks from a vending machine by moving their phones
close to an RFID enabled phone reader. Payment is made through the
person’s mobile phone bill. Consumers pay their bills by simply scanning
the bar code on the bill and using their mobile phones to process payment.

BOX 11.2  Radio Frequency Identification

A Radio Frequency Identification (RFID) tag attached to a product, animal,
or person for the purpose of identification and tracking, makes use of radio
waves. Some tags can be read from several metres away and beyond the line
of sight of the reader.

Interactive advertisements
In an interactive advertisement, customers can scan a bar code in an
advertisement for a product appearing on a TV screen using their mobile
phones. By scanning the bar code, the consumer can order the product by
invoking an internet application.

Catalogue shopping
Mobile phones can be used to place orders for products listed in a
catalogue. For example, a consumer might receive a catalogue by SMS
from a catalogue shopping company. Each product on sale is accompanied
by a unique bar code. By scanning the bar codes, the consumer can buy
products directly from the catalogue shopping company.

11.2  Business-to-Business (B2B) Applications

Business-to-business (B2B) is a form of commerce in which products or

services are sold from a company to its dealers. For example, a company
that manufactures TV sets would normally sell it through a dealer
network rather than selling the product directly to the consumers. Here,
the manufacturer and the dealers are said to be the B2B partners. A few
examples of B2B applications of M-commerce are given below.

Ordering and delivery confirmation

In this application, mobile phones can be used by dealers to order products.
The orders can be sent to the supplier in a standard format. By scanning
the bar code on a product by using the camera of a mobile phone and
specifying the quantity required through a simple application, a dealer can
automatically re-order goods.
Mobile phones can be used to gather information about the status of
consignments during the transport and delivery process. By reading the
 Mobile Commerce   203

bar code on a packet using a mobile device, a truck driver can confirm in
real-time that a consignment has been delivered.

Stock tracking and control

Mobile phones can be used to keep track of the stock in a distributed
inventory system and send updates to a central database. By using a
mobile phone to scan bar codes or RFID tags on products, employees can
update the stock in real time. Mobile phones (as opposed to dedicated
mobile scanners already used in warehouses) are the particularly attractive
tools where the stock is stored in many locations. An example of such an
application is the stock control of apparel items warehoused in the various
department stores.

Supply Chain Management (SCM)

Information about the supply chain processes can be made available via
mobile devices. By scanning an RFID tag using a mobile phone, it is possible
for a manager or anyone in the supply chain to check information about a
product’s state in the supply chain. This kind of accurate information can
help manage the business efficiently.

Mobile inventory management

An interesting new B2B application reported in (Varshney et al., 2000),
envisages a “rolling inventory” consisting of multiple trucks carrying large
amounts of goods. Whenever a store needs certain goods, it locates the
nearest truck to take delivery of the required goods. This reduces the
amount of inventory and cost for both the producers and the retailers. It
also has the potential to drastically reduce the delivery times and help in
just-in-time delivery of goods.

11.3  Structure of Mobile Commerce

In mobile commerce, a content provider implements an application by

providing two sets of programs: client-side and server-side. The client-
side programs run on the microbrowsers installed on the users’ mobile
devices. The server-side programs, performing database access and
computations, reside on the host computer (servers). The architecture of a
mobile commerce framework is shown in Figure 11.1. Below, we explain
the functionalities of the various layers of the architecture of a mobile
commerce framework.

Mobile devices
Hand-held devices essentially present user interfaces to the mobile users.
The users specify their requests using the appropriate interface programs,
204  Fundamentals of Mobile Computing

Figure 11.1  Architecture of a mobile commerce framework.

which are then transmitted to the mobile commerce application on the

Internet. The results obtained from the mobile commerce application are
displayed in suitable formats.

BOX 11.3  Features required of a mobile device to enable mobile commerce

To enable mobile commerce to be used widely, a mobile device should provide
the following facilities:
• Good Internet connectivity
• Ability to display rich content such as images
• Have a good quality camera with auto focus
• Screen should be able to properly display the bar codes
• Ability to read the RFID tags
• MMS (Multimedia Message Service), SMS (Short Message Service)
• Ability to communicate between the mobile device and the supporting
network
• Ability to scan bar codes
• Ability to interact with the Point-of-Sale (PoS) terminals.
Point-of-Sale (PoS) usually means a checkout counter in a shop or supermarket.
More specifically, the point-of-sale often refers to the hardware and software
used for handling customer purchases at the checkout desks. An example of a
PoS terminal is an electronic cash register. Nowadays, the point-of-sale systems
are used in almost every supermarket and are used in many retail stores too.

Mobile middleware
The main purpose of mobile middleware is to seamlessly and transparently
map the Internet content to mobile phones that may sport a wide variety of
operating systems, markup languages, microbrowsers, and protocols. Most
mobile middleware also handle encrypting and decrypting communication
in order to provide secure transactions.

Network
Mobile commerce has become possible mainly because of the availability
of wireless networks. User requests are delivered either to the closest
wireless access point (in a wireless local area network environment) or
 Mobile Commerce   205

to a base station (in a cellular network environment). Wired networks

are optional for a mobile commerce system. However, host computers
(servers) are generally connected to wired networks such as the Internet.
So user requests are routed to these servers using transport and/or security
mechanisms provided by wired networks.

Host computers
Host computers are essentially servers that process and store all the
information needed for mobile commerce applications. Most application
programs used in the mobile commerce are hosted on these. These
applications usually consist of three major components: web servers,
database servers, and application programs and support software. The web
servers help interact with the mobile client. The database servers store data.
The application program is the middleware that implements the business
logic of the mobile commerce application.

11.4  Pros and Cons of M-Commerce

Like any other technology, M-commerce has its own advantages and
disadvantages, which are discussed below.

Advantages
The following are the major advantages of M-commerce:
1. For the business organization, the benefits of using M-commerce
include customer convenience, cost savings, and new business
opportunities.
2. From the customer’s perspective, M-commerce provides the flexibility
of anytime, anywhere shopping using just a lightweight device. The
customer can save substantial time compared to visiting several
stores for identifying the right product at the lowest price.
3. Mobile devices can be highly personalized, thereby providing an
additional level of convenience to the customers. For example, a
repeat order for some items can be placed at the touch of a button.

Disadvantages
The following are the major shortcomings of using M-commerce.
1. Mobile devices do not generally offer graphics or processing power
of a PC. The users are therefore constrained to use small screen and
keyboard and low resolution pictures and videos.
2. The small screens of mobile devices limit the complexity of
applications. For example, the menu choice, and text typing capabi-
lity are severely constrained
206  Fundamentals of Mobile Computing

3. The underlying network imposes several types of restrictions.

For example, the available bandwidth is severely restricted, and
international reach is prohibitively expensive. Therefore, ubiquity
of M-commerce is hard to achieve in practice.

11.5  Mobile Payment Systems

Mobile payments are a natural evolution of E-payment schemes. A mobile

payment (or M-payment) may be defined as any payment instrument where
a mobile device is used to initiate, authorize and confirm an exchange of
financial value in return for goods and services. Mobile devices include
mobile phones, PDAs, and any other device that connects to a mobile
network for making payments. A mobile device can also be used for
payment of bills (especially utilities and insurance premiums) with access
to account-based payment instruments such as electronic funds transfer,
Internet banking payments, direct debit and electronic bill presentment.
An important issue which influences the establishment of the
mobile payment procedure is the technical infrastructure needed on the
customer side. A sophisticated technology may fail if the customer is
not able to handle it with ease. On the other hand, simple procedures
based on simple message exchange via short messaging services (SMS)
may prove more successful. Thus, at present and possibly in the near
future, the important payment solutions will be SMS-based, which
can easily be charged to the mobile phone bill of customers. Some
other procedures may integrate two or more solutions. It is true that
M-payments are still in their infancy. Some important problems
dogging the M-payment schemes are those of security, privacy,
and guarding against frauds. The challenges for providing secure
transactions are many and range from physical theft of a mobile
device which can be subsequently used for fraudulent payments. The
M-payment solutions are still being developed with standards defined
on individual business segments. The other interesting areas related
to M-commerce payment not mentioned in this text are the issues of
standardization and interoperability. In the following section, we discuss
the important types of mobile payment schemes.

11.5.1  Mobile Payment Schemes

Three popular types of M-payment schemes are currently being used:

(a) Bank account based
(b) Credit card based
(c) Micropayment
 Mobile Commerce   207

In each of these approaches, a third party service provider (bank, credit

card company, or telecom company) makes a payment on the customer’s
behalf. An important question that needs to be answered is since the third
party incurs an overhead in making the payment, how would it recover
the cost. First, the service provider may require pre-payment from users,
leading to some financial gain through investment of this fund. A service
provider may charge a small amount as service charge, which can decrease
with increasing customer base.

Bank account based M-payment

In this scheme, the bank account of the customer is linked to his mobile
phone number. When the customer makes an M-payment transaction with
a vendor or in a shopping complex, based on a Bluetooth or wireless LAN
connectivity with the vendor, the bank account of the customer is debited
and the value is credited to the vendor’s account.

BOX 11.4  mChek—a new payment scheme

mChek (Dahlberg et al., 2007) is a new payment system that links a debit
or credit card, or a bank account, to a mobile phone, allowing one to make
payments from the mobile phone. Once registered, the user can pay phone
bills, transfer talktime to a friend’s account, book tickets for flights, movies,
pay water bills, electric bills, etc. from the mobile with the help of simple
instructions. This scheme has a tie-up with Airtel which allows Airtel subscribers
to download mChek application which provides simple graphic interface to
use mChek. There are no charges for downloading as well as using mChek
as of now.

Credit card based M-payment

In the credit card based M-payment, the credit card number is linked to
the mobile phone number of the customer. When the customer makes
an M-payment transaction with a merchant, the credit card is charged
and the value is credited to the merchant’s account. Credit card based
solutions have a limitation, being heavily dependent on the level of
penetration of credit cards in a country. Currently, the penetration level
of credit cards is rather low but is expected to grow substantially in the
coming years.

Micropayment
Micropayment is intended for payment for small purchases such as from
vending machines. The mobile device can communicate with the vending
machine directly using a Bluetooth or wireless LAN connection to negotiate
the payment and then the micropayment is carried out. A customer makes
a call to the number of a service provider where the per call charge is
208  Fundamentals of Mobile Computing

equal to the cost of the vending item. Thus, the micropayment scheme is
implemented through the cooperation of the mobile phone operator and
a third party service provider. This approach has been used for vending
from Coca-Cola machines.

BOX 11.5  Payment settlement solutions

Payment solutions can be categorized on the basis of the payment settlement
methods, which are instant-paid, postpaid, prepaid, or a combination of these. In
the prepaid solution, customers buy a smart card where the amount equivalent
is stored and then they can pay for goods or services desired. Subscription to
services can also be considered as the prepaid type of payment. The prepaid
type of solution allows privacy to users since at no point of the process is it
required to disclose any personal data. The instant-paid solution is that in which
payment settlement is done as soon as users confirm the payment as in direct
debiting systems. In the postpaid solution, customers pay for goods or services
later. Payment by credit card and phone bill is an example of this solution.

11.6  Security Issues

M-commerce is anticipated to introduce new security and privacy risks

beyond those currently found in E-commerce systems (Ghosh and
Swaminatha, 2001). Users of mobile devices can be difficult to trace because
of roaming of the users. Also, the mobile devices go on-line and off-line
frequently. Thus, attacks would be very difficult to trace.
Another risk unique to the mobile devices is the risk of loss or theft. A
mobile device that is stolen or has fallen into wrong hands can cause frauds
that are difficult to track and prevent. A major problem in this regard is
the lack of any satisfactory mechanism to authenticate a particular user.

SUMMARY

M-commerce denotes ubiquitously carrying out any activity associated

with buying or selling of goods. M-commerce gives added flexibility, cost
advantages, and conveniences to both the buyer and the seller of goods.
For the sellers, M-commerce provides significant business benefits and
market reach. The customers benefit by getting the best deal available.
M-commerce even gives the traditional media (newspapers, magazines,
TV) the opportunity to use the advertisements appearing in their media
in innovative ways such as automated searching of recent advertisements
by customers in a shopping complex. The wireless service providers can
also see increased traffic from M-commerce. Considering these benefits to
all the parties concerned, M-commerce is poised to take off in a big way.
 Mobile Commerce   209

FURTHER READINGS

Dahlberg, T., et al., “Past, present and future of mobile payments research:
a literature review,” Electronic Commerce Research and Applications,
doi:10.1016/j.elerap, 2007.02.001.
Ghosh K. Anup, and T. Swaminatha, “Software Security and Privacy
Risks in Mobile E-Commerce,” Communications of the ACM, pp. 51–54,
February 2001, http://main.mchek.com.
Varshney, U., R.J. Vetter, and R. Kalakota, “Mobile commerce: a new
frontier”, IEEE Computer, Vol. 33, pp. 32–38, October 2000.
Zheng, X. and D. Chen, “Study of mobile payments systems,” IEEE
International Conference on E-Commerce, CEC 2003, pp. 24–27, Digital
Object Identifier 10.1109/COEC.2003.1210227, June 2003.

EXERCISES
1. What do you understand by the mobile payment system? Briefly
explain an application where mobile payment may be useful. Explain
the different payment systems that are available.
2. What do you mean by the 4 Ps of commerce? Explain the different
forms of commerce that are obtained by varying the interpretation
of the Ps.
3. What do you understand by M-commerce? What are the advantages
and disadvantages of M-commerce.
4. Explain any two applications of M-commerce.
5. What do you mean by B2B and B2C commerce? Give examples of
M-commerce for these two categories of commerce.
6. What is RFID? Briefly explain the principle of its working. Explain
a few applications in which RFID is useful.
7. What is micropayment in M-commerce? How is micropayment
achieved?
 Glossary

Access Point:  It is connected to a wired network through cables and

provides the connectivity between the wireless device and the wired
network.
Ad Hoc On-Demand Distance Vector (AODV):  AODV has the basic
route-discovery and route-maintenance like the DSR protocol and uses
the hop-by-hop routing, sequence numbers and beacons.
Adaptation:  Adaptation implies the ability of a system to adjust to
bandwidth fluctuation without inconveniencing the user.
Agent Advertisement:  Generally the foreign and the home agents advertise
their presence through periodic agent advertisement messages. An
agent advertisement message, lists one or more care-of-addresses and
a flag indicating whether it is a home agent or a foreign agent. Agent
advertisement is a popularly used method in agent discovery.
Agent Discovery:  During call establishment it is necessary for a mobile
node to determine its foreign agent. This task is referred to an agent
discovery. The two discovery methods used are:
1. Agent advertisement and 2. Agent solicitation.
Agent Solicitation:  In case a mobile node (MN) does not receive any
care-of-address COA, then the MN should send an agent solicitation
message.
Amplifier:  An amplifier amplifies the strength (usually voltage) of a
signal.
AMPS:  The first generation (1G) of the cellular system was designed
in the late 1960s and was deployed commercially in the early 1980s.
The first commercial 1G system in the United States was known as
Advanced Mobile Phone System (AMPS). This was a completely analog
system.
211
212  Glossary

Antenna:  The functions of an antenna are to convert electrical signals

from a transmitter to a propagating electromagnetic RF wave, or
conversely, to convert a propagating RF wave to electrical signals in
a receiver. In a transceiver, a transmitter and a receiver are co-located
for full-duplex communications, the same antenna is usually used to
both transmit and receive.
Application Tier:  Application tier is also known as the middle tier. This
layer co-ordinates the application, processes the commands, makes
logical decision evaluations and performs the calculations.
ARP (Address Resolution Protocol):  The ARP protocol is used by IP to
find the hardware (physical) address of a computer network card based
on the IP address.
Atomicity:  Either all operations of a transaction are reflected on the
database or none at all.
Atomicity Relaxation:  In this modified scheme, an MH is allowed to
submit ‘pieces’ of transactions from different cells according to their
movements. This approach weakens the atomicity property and requires
the ability of breaking a transaction so that the sub-transactions can
be concurrently executed and interleaved with the sub-transactions of
other transactions while guaranteeing other ACID properties.
Automated Local Evaluation in Real Time (ALERT):  ALERT is an
automated flood alarming system that is probably the first well-
known real wireless sensor network deployment. ALERT provides the
important real-time rainfall and water level information to evaluate the
possibility of potential flooding. ALERT sensor networks are usually
equipped with meteorological sensors. Currently, ALERT is deployed
across most of the western United States such as California and Arizona.
B2B Ordering and Delivery Confirmation:  Mobile phones can be used
by dealers to order products. The orders can be sent to the supplier
in a standard format. By scanning the bar code on a product with
the camera of a mobile phone and using a simple application to state
the quantity required, the owner of a small shop can automatically
re-order goods.
B2B Supply Chain Management (SCM):  Information about the product’s
state in a supply chain process can be checked via mobile devices in
order to manage the business efficiently.
B2B Tracking:  Mobile phones can be used to keep track of the stock in a
distributed inventory system and send updates to a central database.
By using a mobile phone to scan bar codes or RFID tags on products,
employees can update the stock in real time.
B2C Advertising:  Advertisements are sent directly to mobile phones. For
example, suppose a consumer in a shop is fascinated by a new electronic
product and wishes to buy it but only after getting more details about
 Glossary   213

it. For this purpose, he can view all the relevant advertisements for the
product by taking the picture of the bar code using his mobile device.
B2C Catalogue Shopping:  Mobile phones can be used to place orders
for products listed in a catalogue. For example, a consumer receives a
catalogue by SMS from a catalogue shopping company. Each product
on sale is accompanied by a unique bar code. By scanning the bar
codes, the consumer can buy products directly from the catalogue
shopping company in a similar way to accessing a web-based catalogue
on his mobile phone.
B2C Information About a Product:  Consumers can access additional
information about products through their mobile phones.
B2C Interactive Advertisement:  An interactive advertisement allows
customers to scan a bar code appearing on a TV screen using their
mobile phone.
B2C Loyalty and Payment Services:  Mobile phones can replace the
physical loyalty cards. Having signed up for a supermarket loyalty
scheme, a unique bar code is sent to a consumer’s mobile phone. After
shopping at the same supermarket, the consumer shows the bar code
at the cash counter and accumulates points based on the total amount
spent.
B2C Mobile Ticketing:  Mobile phones can be used to purchase and
redeem tickets. A traveller can buy movie tickets online.
B2C Shopping:  Consumers can use their mobile phones to get a
comparative pricing analysis of a product at different stores and also
the prices of the related products.
Bandwidth Constrained, Variable Capacity Links:  Wireless links have
significantly lower capacity than their wired counterparts. Also, due to
issues such as multiple access, fading, noise and interference conditions,
the bandwidth of the wireless links can change arbitrarily with time.
Black Hole:  In this type of attack, a node can set up a route to some
destination via itself, and when the actual data packets are received
they are simply dropped. This forms a black hole, where data enter
but never leave.
Blacklist:  Some ad hoc routing protocols try to tackle the security
problem by keeping a list of perceived malicious nodes. Each node
has a blacklist of, what it thinks, bad nodes and thereby avoids using
them when setting up routing paths.
Bluetooth Baseband:  It is concerned with connection establishment and
other tasks such as packet format, timings, etc.
Bluetooth Link Manager Protocol (LMP):  It is responsible for link setup
between the Bluetooth devices. It also includes some security issues
such as (i) authentication and (ii) negotiation and control of baseband
packet sizes.
214  Glossary

Bluetooth Logical Link Control and Adaptation Protocol (L2CAP):  It

enables linking the baseband layer to the upper-layer protocols. This
L2CAP can provide (i) connectionless services and (ii) connection-
oriented services.
Bluetooth Radio:  It specifies the details of wireless support that include
(i) frequency (ii) use of frequency hopping (iii) modulation scheme and
(iv) transmit power.
Bluetooth Radio Frequency Communication (RFCOMM):  It is a cable
replacement protocol used in Bluetooth specification. RFCOMM
represents a virtual serial port, designed to replace cable technologies.
With minimum changes in existing devices, RFCOMM replaces the
serial port cables and provides data transport in Bluetooth.
Bluetooth Services Discovery Protocol (SDP):  It can supervise and enable
connection between two or more Bluetooth devices for a specific service
application. SDP includes information such as (i) device information
(ii) services and (iii) service characteristics.
Bluetooth Technology:  Bluetooth technology enables users to easily
connect to a wide range of computing and telecommunications
devices, without the need to buy, carry, or connect cables. It provides
opportunities for rapid ad hoc connections, and the possibility of
automatic, transparent connections between devices. It eliminates
the need to purchase additional or proprietary cabling and configure
exercises to connect individual devices.
Bluetooth TelCP:  Bluetooth also specifies the telephony control protocol. It
defines the mobility management functions so that groups of telephony
control specification (TCS) devices can be handled. TCS is a bit-oriented
protocol that defines call control signalling. It helps to establish speech
and data calls between the Bluetooth devices.
BOOTP (Boot Protocol):  BOOTP is a network protocol used by a network
client to obtain an IP address from a configuration server. BOOTP is
usually used by a diskless computer when a computer is starting up.
A BOOTP configuration server assigns an IP address to each client
from a pool of addresses.
Bridge:  It is used for connecting two LANs that may be in two different
buildings or on two separate floors within the same building.
Broadcast:  In this type of transmission, the message is sent to all the
network nodes.
Bus Architectures:  In bus-based architectures, nodes are connected
to the network cable using T-shaped network interface connectors.
Terminating points are placed at each end of the network cable.
Business-to-Business (B2B):  Business-to-business (B2B) is a form of
commerce in which products or services are sold from a company to
dealers.
 Glossary   215

Business-to-Consumer (B2C):  Business-to-Consumer (B2C) is a form of

commerce in which products or services are sold from a business firm
to a consumer.
Campus Wireless LANs:  Several organizations, hotels, retail outlets,
warehouses, factories, research centres, and educational institutions,
among many others, are recognizing the value of flexibility and
connectivity provided by wireless LANs.
Care-of-Address (COA):  It is the address that is used to identify the
present location of a foreign agent. The packets sent to the MN are
delivered to COA.
Centralized Environment:  This is the oldest data-processing system
in which a large program manages both the tasks of initiating a
transaction by interacting with users and processing the transaction.
In this environment, the overhead associated with the processing of
a transaction is very low as a single user executes the transactions.
Recovery is also a trivial issue here, since rollback recoveries do not arise.
Client-Server Environment:  In this environment, the execution of a
transaction and transaction initiations are done by a server and a client,
respectively. The server in this environment is typically concurrent.
That is, many clients can be connected to the server at the same time
to submit transactions simultaneously.
Code Division Multiple Access (CDMA):  CDMA is an access method
in which multiple users are allotted different codes of sequences of
0s or 1s to access the same channel. A special coding scheme is used
that allows multiple users to be multiplexed over the same physical
channel.
Co-located COA:  When the mobile node (MN) has acquired a temporary
IP address, then that address acts as COA.
Configuration Layer:  The configuration layer defines the minimum set
of Java Virtual Machine features and Java class libraries available on
a particular category of devices. In a way, a configuration defines the
commonality of the Java platform features and libraries that developers
can assume to be available on all devices belonging to a particular
category. This layer is less visible to users, but is very important to
profile implementers.
Connected Device Configuration (CDC) for Plug-in Devices:  CDC
devices use a 32-bit architecture, have at least 2 MB of memory
available, and implement a complete functional JVM. CDC devices
include digital set-top boxes, home appliances, navigation systems,
point-of-sale terminals, and smartphones.
Connected Limited Device Configuration (CLDC) for Hand-held
Devices:  CLDC is aimed at the low end of the consumer electronics
range and designed for 16-bit or 32-bit small computing devices with
216  Glossary

limited memory. These devices usually have between 160 KB and

512 KB of available memory.
Consistency Relaxation:  Under this approach, the database is logically
partitioned into ‘clusters’, based on some criteria such as certain
semantic properties or location proximity. The data in the same
cluster must be strictly consistent, whereas a ‘bounded’ degree of
inconsistency is tolerated amongst clusters, according to a relaxed
definition of consistency. In this approach, a mobile host can download
one cluster before disconnection and execute transactions on this data
by preserving the ACID properties and after reconnection it can be
copied to the original database if the inconsistencies are below some
agreed value.
Consistency:  A transaction should transform the database from one
consistent state to another consistent state.
Correspondent Node (CN):  The home agent is a router on the home
network serving as the anchor point for communication with the
mobile node. It tunnells packets from a device on the Internet, called
a correspondent node, to the roaming mobile node.
Data Replication:  Data replication is the process of maintaining a defined
set of data in more than one location. It involves copying the designated
changes from one location (source or master) to another (target or
remote), and synchronizing the data in both the locations.
Data Tier:  The data tier consists of database management system and
data store.
Data-centric:  In many applications of sensor networks, it is not possible
to assign a global identifier to each node due to the sheer number of
nodes deployed. Lack of such global identifiers along with random
deployment of sensor nodes, makes it hard to select a specific set of
sensor nodes to be queried. Therefore, data is usually transmitted
from every sensor node within the deployment region with significant
redundancy.
Denial of Service (DoS):  It is possible for a malicious node to continuously
transmit and to monopolise the channel use and cause other nodes to
wait endlessly.
Destination-Sequenced Distance-Vector Routing (DSDV) Protocol:  The
Destination-Sequenced Distance-Vector Routing (DSDV) protocol is a
table-driven algorithm based on the classical Bellmon–Ford routing
algorithm. An improvement made here is the avoidance of routing
loops. Each node in an ad hoc network maintains a routing table
in which all of the possible destinations within the non-partitioned
network and the number of routing hops to each destination are
recorded. Hence, routing information is always made readily available,
regardless of whether the source node requires a route or not.
 Glossary   217

DHCP Automatic Allocation:  In automatic allocation, DHCP assigns a

permanent IP address to a particular client.
DHCP Dynamic Allocation:  In dynamic allocation, DHCP assigns an IP
address to a client for a specific period of time.
DHCP Manual Allocation:  In manual allocation, a client’s IP address is
assigned by the network administrator, where the DHCP is used to
inform the address assigned to clients.
Directed Diffusion (DD):  In this protocol, each sensor node names its
data with one or more attributes. A destination node sends interests for
data, based on these attributes. Interests are flooded over the network.
When a node receives an interest from a neighbour, it sends data to the
neighbour. Each node only knows the neighbour from whom it got the
interest. It is possible that each node would receive the same interest
from more than one neighbour. In this way, multiple paths can be set
up from the source node to the destination node.
Disconnections:  A fairly complicated issue in the operation of a mobile
database arises due to the temporary disconnections of mobile users
arising out of attempts to save power.
Discovering the Care-of-Address:  The discovery of the care-of-address
consists of three important steps. (1) Mobile agents advertise their
presence by periodically broadcasting through the Agent Advertisement
messages. (2) The mobile node receiving the Agent Advertisement
message observes whether the message is from its own home agent and
determines whether it is on the home network or a foreign network.
(3) If a mobile node does not wish to wait for the periodic advertisement,
it can send out Agent Solicitation messages that will be responded by
a mobility agent.
Distance Vector Protocols:  The distance vector algorithm is based on
the Bellman–Ford algorithm. Each node advertises its entire routing
table to its immediate neighbours only. As a node receives the routing
information of its neighbouring nodes, it updates its own routing table
by looking at its neighbours’ routing table.
Distributed Coordination Function (DCF):  It is a distributed function. It
can be used with both the infrastructure and the ad hoc (centralized)
configurations.
Distributed Environment:  In this environment, data is organized in a
distributed fashion over a network. That is, either a transaction can
get fully executed at any arbitrary node, or some parts of transaction
can be executed at a node. Atomicity and durability of a transaction
become critical issues in this environment.
DNS (Domain Name Server):  Aliases used for IP addresses are called
domain names. When we address a website (i.e specify its URL)
the domain name is translated to the corresponding IP address by a
Domain Name Server.
218  Glossary

Dropping Routing Traffic:  It is essential in an ad hoc network that all

nodes participate in the routing process. However, a node may act
selfishly and process only the routing information that are related to
itself in order to conserve energy. This behaviour/attack can create
network instability or can even segment the network.
Durability:  Changes made to the database by a committed transaction
must persist even after any failure.
Durability Relaxation:  Durability of committed transactions is mainly
affected by the possibility of MHs autonomously operating on data.
A disconnected MH can only commit a transaction locally if this
transaction does not conflict with other transactions executed on
the same host while the host is disconnected. On reconnection, the
transactions are globally executed on different MHs.
Dynamic Host Configuration Protocol (DHCP):  DHCP was developed
based on bootstrap protocol (BOOTP). DHCP provides several types
of information to the user (client) including its IP address. To manage
dynamic configuration information and dynamic IP addresses,
IETF standardized an extension to BOOTP known as dynamic host
configuration protocol (DHCP).
Dynamic Source Routing (DSR):  Dynamic Source Routing (DSR) is a
source initiated, on-demand routing protocol for ad hoc networks.
It uses source routing, a technique in which the sender of a packet
determines the complete sequence of nodes through which a packet
has to travel.
Dynamic Topologies:  Since nodes are free to move arbitrarily, the
network topology may change unpredictably.
Filters:  Filters are the key components of all wireless transmitters and
receivers. They are used to reject interfering signals lying outside the
operating band of receivers and transmitters. They also reject unwanted
noise generated by amplifiers.
Foreign Agent COA:  The COA is an IP address of a foreign agent (FA).
Foreign Agent:  The foreign agent is a router in a foreign network that
functions as the point of attachment for a mobile node when it roams
to the foreign network. The packets from the home agent are sent to
the foreign node which delivers it to the mobile node.
Foreign Network:  The foreign network is the current subnet to which the
mobile node is visiting. It is different from home network. A foreign
network is the network in which a mobile node is operating when
away from its home network.
Freeze-TCP:  The basic idea in this scheme is to “freeze” the TCP senders’
streams a little before a disconnection is to occur.
 Glossary   219

Frequency Division Multiple Access (FDMA):  For systems using

Frequency Division Multiple Access (FDMA), the available bandwidth
(frequency range) is subdivided into a number of narrower band
channels. Each user is allocated a unique frequency band to transmit
and receive signals.
FTP (File Transfer Protocol):  The FTP protocol is used to transfer files
between computers.
General Packet Radio Service (GPRS):  GPRS when integrated with GSM,
significantly improves and simplifies Internet access. It transfers data
packets from GSM mobile stations to external packet data networks
(PDNs). Packets can be directly routed from the GPRS mobile stations
to packet-switched networks, making it easy to connect to the Internet.
Geographic Adaptive Fidelity (GAF):  GAF is an energy-aware location-
based routing algorithm, designed primarily for the mobile ad hoc
networks but is applicable to sensor networks as well. The network
area is first divided into fixed zones forming a virtual grid. Inside
each zone, nodes collaborate with each other to play different roles.
For example, nodes elect one sensor node to stay awake for a certain
period of time while the others go to sleep mode.
Geographic and Energy Aware Routing (GEAR):  This protocol is based
on the knowledge of node’s own as well as its neighbours’ locations
and energy information. It conserves energy by minimizing the number
of interests in DD by only considering a certain region rather than
sending the interests to the whole network.
Global System for Mobile Communication (GSM):  Possibly, the most
successful digital mobile system in current usage is the GSM . The
main goal of GSM was to provide a mobile phone system that allows
data services and provides voice services compatible to the previous
generation systems.
GPRS Architecture:  GPRS architecture introduces two new network
elements, called the GPRS Support Node (GSN) and the Gateway
GPRS Support Node (GGSN).
GPRS Services:  GPRS offers end-to-end packet-switched data transfer
services which are mainly categorized into two types:  (1) Point-To-
Point (PTP) service and (2) Point-To-Multipoint (PTM) service.
GPRS:  Global Packet Radio System (GPRS) is an extension of GSM and is
considered to be 2.5 generation technology. As indicated by the name,
it is based on packet switching, compared to circuit switching in the
previous generation technologies. An important advantage of GPRS
is that it allows users to remain connected to the Internet without
incurring additional charge and supports multimedia capabilities
including graphics and video viewing.
220  Glossary

Great Duck Island (GDI) System:  This is an example of Habitat

Monitoring Application. The researchers from Intel Research laboratory
deployed a mote-based sensor network on Great Duck Island, Maine,
USA to monitor the behaviour of storm petrel (small sea birds with
tube shaped nostrils).
GSM Anonymity:  A GSM network protects against someone tracking
the location of the user or identifying calls made to or from the user
by eavesdropping on the radio path.
GSM Authentication Centre (AuC):  AuC is related to the Home Location
Register (HLR). The AuC stores information that is concerned with
security features, i.e. user authentication and encryption.
GSM Authentication:  The purpose of authentication is to protect the
network against any unauthorized use. It also enables the protection
of the GSM subscribers by denying the possibility for intruders to
impersonate authorized users. The GSM network operator can verify
the identity of the subscriber making it infeasible to clone someone
else’s mobile phone.
GSM Base Station Controller (BSC):  The BTSs are managed by the BSC.
It reserves radio frequencies, handles the handover from one BTS to
another within the BSS. The BSC also multiplexes the radio channels
onto the fixed network connections at the interface.
GSM Base Station Subsystem (BSS):  A GSM network comprises many
BSSs, each BSS contains a BSC and several BTSs. Each BSS is controlled
by the Base Station Controller (BSC). The BSS performs all functions
necessary to maintain radio connections to an MS, and coding/
decoding of voice.
GSM Base Transceiver Stations (BTS):  A BTS comprises all radio
equipment such as antenna, signal processing, amplifiers necessary
for radio transmission. A BTS can form a radio cell and is connected
to an MS via an interface.
GSM Bearer Services:  Bearer services can either be connection-oriented or
packet-switched. They comprise all services that enable the transparent
transmission of data to the network. These services are implemented
on the lower-three layers of the OSI reference model.
GSM Confidentiality:  A GSM network protects voice, data and sensitive
signalling information (e.g. dialed digits) against eavesdropping on the
radio path.
GSM Equipment Identity Register (EIR):  It contains mobile equipment
identity information which helps to block calls from stolen, unauthorized,
or defective mobile stations.
GSM Home Location Register (HLR):  It stores information that is specific
to each subscriber including the user’s current location.
 Glossary   221

GSM Mobile Services Switching Centre:  Mobile Services Switching

Centres (MSCs) form the fixed backbone network of a GSM system.
MSCs set up connections to other MSCs and to the BSCs and also
connect to Public Data Network (PDN). MSCs are responsible for
connection setup, connection release and handover of connections to
other MSCs. They also perform supplementary services such as call
forwarding, multiparty calls, etc.
GSM Mobile Station (MS):  The mobile stations or cell phones contain
two major components: the Subscriber Identity Module (SIM), which
is a removable smart card, and the Mobile devices.
GSM Network and Switching Subsystem:  This subsystem forms the
heart of the GSM system. It connects the wireless networks with the
standard public networks and performs and supports usages charging,
accounting, and handles roaming of users between different countries.
The NSS consists of the switching centre and several types of database.
GSM Operation and Maintenance Centre (OMC):  OMC supervises all
other network entities. Its functions are traffic monitoring, subscribers
management, security management and accounting billing.
GSM Operation Subsystem:  The operation subsystem contains all the
functions necessary for network operation and maintenance.
GSM Radio Subsystems:  This subsystem comprises all the radio specific
entities, i.e. the mobile stations, the base station subsystems, the base
transceiver station and the base station controller.
GSM Security:  Security in GSM is broadly supported at three levels:
operator’s level, customer’s level and system level. These three levels,
oversee aspects such as correct billing to the customer, preventing
fraud, protecting services, privacy, and anonymity.
GSM System Architecture:  A GSM system consist of three main
subsystems:  Radio Subsystem (RSS), Networking and Switch
Subsystem (NSS) and Operation Subsystem (OSS).
GSM Tele Services:  GSM provides voice-oriented tele services as well
as non-voice tele services. The voice-oriented tele services constitute
telephony and the non-voice tele services constitute SMS and FAX.
GSM Visitor Location Register (VLR):  It is associated with one or more
MSCs and it contains information relating to those subscribers that
are currently registered within the MSC area(s) of its associated MSC.
VLR is used to store the information of users who are currently in the
area of control.
Hierarchical Routing:  These protocols are also known as cluster-based
routing. In these protocols, different nodes can play different roles in
the network based on their remaining battery powers. Higher-energy
nodes are used to process and send the information, while the low-
energy nodes are used to perform sensing in the proximity of the target.
222  Glossary

Home Address:  The home address of a mobile device is the IP address

assigned to the device within its home network. The IP address on the
current network is known as the home address.
Home Agent (HA):  It is located in home network and it provides several
services for the mobile node (MN). HA maintains a location registry.
The location registry keeps track of the node location by the current
care-of-address of the MN.
Home Network:  The home network of a mobile device is the network
within which the device receives its identifying IP address (home
address). In other words, a home network is a subnet to which a mobile
node belongs to as per its assigned IP address.
HTTP (Hyper Text Transfer Protocol):  The HTTP protocol is used for
communication between a web server and a web browser.
ICMP (Internet Control Message Protocol):  The ICMP protocol is used
for error handling in the network.
IEEE MAC Standard:  The IEEE 802.11 is the most widely used standard
for WLANs today. The IEEE 802.11 standard defines the functional
aspects of the medium access control (MAC) sublayer. The nodes in
the IEEE 802.11 WLAN can be grouped in one of two configurations:
infrastructure (with Access Point) and ad hoc (without Access Point).
IEEE:  The IEEE is a non-profit, technical professional association of
more than 377,000 individual members in 150 countries. The IEEE
acts as a standards body, since in networking, multiple devices that
are possibly heterogeneous and manufactured by different vendors
need to communicate. The formation of suitable standards is especially
important in networking. The IEEE proposes standards for new
technologies and maintains the old standards. The IEEE proposed the
802.11 standard for wireless networking. The standard 802.11 is the
generic name of a family of standards for wireless networking. The
numbering system for 802.11 comes from the IEEE, who uses “802”
for many networking standards like Ethernet (802.3).
IGMP (Internet Group Management Protocol):  IGMP is a protocol
through which hosts exchange information with their local routers.
The routers use the IGMP to check whether a known group members
are active or not.
Indirect TCP:  The Indirect-TCP (I-TCP) protocol suggests that any TCP
connection from a mobile host to a machine on the fixed network
should be split into two separate connections:  one TCP connection
between the fixed host and the base station, and the other between
the base station and the mobile host.
Integrity of Database:  To ensure the integrity of a database, a DBMS
should maintain the following constraints:
 Glossary   223

Intrusion Detection in MANETs:  Recognizing and responding to

malicious activities, is known as intrusion detection. Intrusion detection
is used as a process, which involves technology, people and tools.
Intrusion detection is an approach that is complementary with respect
to mainstream approaches to security such as access control and
cryptography.
IP (Internet Protocol):  IP is responsible for delivering data packets across
the Internetwork based on their addresses.
IP Address:  Each computer must have an IP address before it can
meaningfully connect to the Internet. That is, each IP packet must
have a destination IP address before it can be sent to another computer.
IP Router:  The IP router is responsible for “routing” the packet to its
destination, directly or via other routers, based on its IP address.
Isolation Relaxation:  Some transaction models have been devised for
mobile environments in which the isolation property is not guaranteed.
That is, intermediate results of a transaction can be observed by other
transactions. This is usually a side effect of the relaxation of other
ACID properties.
Isolation:  The effect of a transaction should be such that it appears to be
executed in isolation, That is, its intermediate results must not be seen
by other transactions.
J2ME:  The Java2 Micro Edition (J2ME) is meant for tiny devices such as
mobile phones, TV set-top boxes, pagers, PDAs, etc.
Java Virtual Machine Layer:  This layer is an implementation of a Java
Virtual Machine that is customized for a particular device’s host
operating system and supports a particular J2ME configuration.
Lack of Physical Boundary:  Each mobile node functions as a router and
forwards packets from other nodes. As a result, network boundaries
become blurred. The distinction between nodes, internal and external
to a network, becomes meaningless, making it difficult to deploy
firewalls, monitor traffic, etc.
LAN Architecture:  Two major LAN architectures are being used, the
bus architecture and the ring architecture. These two architectures use
different access control techniques.
Limited Computational Capabilities:  Typically, nodes in ad hoc networks
have very limited computational capability. Therefore, handling
public-key cryptography during normal operations would be a severe
overhead.
Limited Physical Security:  Mobile networks are prone to many more
types of security threats than those faced by fixed cable networks,
mainly due to the wireless transmissions and collaborative routing.
There are increased possibilities of eavesdropping, spoofing, denial-
224  Glossary

of-service attacks in mobile networks. Also, nodes are vulnerable to

capture and compromise.
Limited Power Supply:  Since nodes normally rely on battery power, an
intruder can exhaust batteries by causing unnecessary transmissions or
excessive computations to be carried out by nodes.
Link State Protocols (LSP):  In this approach, each node independently
calculates the next best logical path from it to every possible destination
in the network. This is done periodically or whenever a change is
detected in one of the outgoing links.
Location Awareness:  A hand-held device equipped with the global
positioning system (GPS) can provide information about the current
location of a user.
Location-based Routing:  In these protocols, the nodes are addressed by
their location. Distances to next neighbouring nodes can be estimated
by the received signal strengths or by GPS receivers. In this type of
protocols, positions of sensor nodes are exploited to route data to the
network. The distance between neighbouring nodes can be estimated
on the basis of the incoming signal strength. Relative coordinates of
neighbouring nodes can be obtained by exchanging such information
between neighbours. If there is no activity, nodes go to sleep to save
energy. To maximize energy savings, it is necessary to have as many
sleeping nodes in the network as possible.
Low Energy Adaptive Clustering Hierarchy (LEACH):  Low energy
adaptive clustering hierarchy proposed by Heinzelman is a well known
routing protocol in WSN based on hierarchical structure. This protocol
divides the network into clusters (sections). In each cluster there is
an elected sensor node to act as head of the cluster, identified as
cluster head. The cluster head’s task is to manage communication
among member nodes of the cluster, do data processing, and relay the
processed sensed data to the base station.
MAC Protocol:  When multiple nodes contend to access, i.e. transmit on
a shared medium at the same time, the medium access control (MAC)
protocol decides which host would be allowed to transmit.
MAC:  MAC is a sublayer of link layer and directly invokes the physical
layer protocol.
MACA:  MACA stands for Multiple Access Collision Avoidance. MACA
is designed to solve the hidden/exposed terminal problems mainly by
regulating the transmitter power.
Maintaining Logs at the Current MSS:  In this approach, logs are
maintained at the current MSS for all transactions submitted from its
cell. When an MH moves out of the cell, all logs are transferred to the
new MSS. This approach generates a high network traffic due to the
transfer of logs. Also, it introduces additional complications such as
 Glossary   225

how much resources should be allotted to a mobile host, particularly

to foreign ones, and what to do with the logs if an MH fails.
Maintaining Logs at the Mobile Host:  This may be difficult due to lack
of sufficient resources such as memory and processing power. As a
result, storing the logs at the mobile host and initiating the recovery in
case of failure would be an unacceptable overhead. Also, as the failure
rate of a mobile host is high, this approach is further complicated and
not considered satisfactory.
Maintaining Logs at the Mobile Host’s Base MSS:  This approach
incurs communication overhead because each transaction log must be
updated, introducing one round trip delay, which is undesirable.
MANET Hybrid Routing Protocol:  Hybrid routing protocols have
characteristics of both proactive and reactive protocols. These protocols
try to combine the good features of both the protocols.
MANET Proactive Protocols (Table Driven):  In a table-driven (proactive)
routing protocol, each node maintains routes to every other node. The
routing information is usually kept in a number of different tables.
These tables are periodically updated if the network topology changes.
MANET Reactive Protocol (On-Demand):  Reactive routing is also
known as on-demand routing. These routes were designed to reduce
the overheads associated with proactive protocols by maintaining
information for active routes only. Obviously, they do not have to
maintain or constantly update their route tables with the latest route
topology.
MANET Unicast Routing Protocols:  Unicast routing protocols in MANET
are classified into proactive (table-driven), reactive (on-demand) and
hybrid types. This classification is based upon how they respond to
any change to the network topology.
Mesh-based Protocol:  Mesh-based schemes establish a mesh of paths that
connect the sources and destinations. They are more resilient to link
failures as well as to mobility.
MIME (Multi-purpose Internet Mail Extensions):  The MIME protocol
lets SMTP encode multimedia files such as voice, picture, and binary
data across TCP/IP networks. This helps e-mails to include picture,
voice, and binary data files.
Mixers:  A mixer is typically used for frequency conversion in transmitters
and receivers.
Mobile Commerce:  Mobile commerce, essentially involves buying and
selling of commodities, services, or information on the Internet using
mobile hand-held devices. M-commerce offers consumers significant
convenience and flexibility and is playing an increasingly important
role in modern times.
226  Glossary

Mobile Computing:  It is widely described as the ability to compute and

communicate while on the move. Mobile computing encompasses two
separate and distinct concepts: mobility and computing. Computing
denotes the capability to automatically carry out certain processing.
Mobility, on the other hand, provides the capability to change location
while communicating and computing.
Mobile Equipment (ME):  ME is a physical device with unique identifiers
called IMEI (International Mobile Identification Number).
Mobile Information Device Profile (MIDP) Layer:  This layer is a set of
Java APIs that address issues such as user interface, persistence storage,
and networking.
Mobile Internet Protocol (Mobile IP):  Mobile Internet Protocol (Mobile
IP) has been proposed by the Internet Engineering Task Force (IETF).
The mobile IP allows mobile computers to stay connected to the
Internet regardless of their location and without changing their IP
address. In other words, Mobile IP is a standard protocol that builds
on the Internet Protocol by making mobility transparent to applications
and higher level protocols such as TCP.
Mobile Middleware:  The main purpose of mobile middleware is to
seamlessly and transparently map the Internet content to mobile phones
that may sport a wide variety of operating systems, markup languages,
microbrowsers, and protocols. Most mobile middleware also handle
encrypting and decrypting communication in order to provide some
level of security for transactions.
Mobile Node (MN):  A mobile node is a hand-held equipment with
roaming capabilities. It can be a cell phone, a personal digital assistant,
or a laptop.
Mobile Payment Schemes:  Mainly, there are three types of schemes
available for M-payment: (a) Bank account based (b) Credit card based
and (c) Telecommunication company billing based.
Mobile Payment System:  A mobile payment (or M-payment) may be
defined as any payment instrument where a mobile device is used to
initiate, authorize and confirm an exchange of financial value in return
for goods and services.
Mobile TCP:  The design of this protocol is based on the observation that,
in mobile networks, users will be plagued by frequent disconnection
events (caused either by signal fades, lack of bandwidth or by handoff)
that must be explicitly handled by the protocol. Unlike I-TCP, M-TCP
is another split-connection approach that breaks up a TCP connection
between a FH and a MH into two parts: one between the FH and the
BS, and the other between the BS and the MH.
Mote:  The name mote was given by the scientists from the University
of California Berkeley, while working on a project known as Smart
 Glossary   227

Dust. It was a project funded by the Defense Advanced Research

Projects Agency’s (DARPA) Network Embedded Software Technology
program. The aim of the Smart Dust project was to shrink the size of
the sensor nodes (motes) to that of dusts.
Movement of Users:  Movement of users introduces several complications
as far as database operations are concerned. The complications arising
due to user mobility must be satisfactorily addressed.
Multicast:  In this type of transmission the message is sent to a selected
subset of the network nodes.
Multipath-based Routing:  These protocols offer fault tolerance by having
at least one alternative path (from source to destination). However this
increases the energy consumption and the traffic.
Negotiation-based Routing:  It uses negotiation as the basis for cooperation
between the competing entities, for some cases of routing between two
neighbouring ISPs. Inter-domain routing is often driven by self-interest
and based on a limited view of the internet work, which hurts the
stability and efficiency of routing.
Network Size and Node Density:  Network size refers to the geographical
coverage area that could be covered by the network. The number of
nodes present per unit geographical area represents network density.
Network size and node density are the important parameters that
primarily determine the type of routing protocol that would be suitable.
Network Topology:  User mobility affects the network topology. Also
nodes can become inoperative due to discharged batteries leading to
rapid changes to topology.
Operating System (OS):  The operating system (OS) is a system software
that manages the efficient utilization of the resources of a computer by
multiple tasks and provides a user interface communication with base
station to the underlying hardware.
Operational Environment:  The operational environment of a mobile
network refers to a terrain of operations. Common operational
environments include urban, rural and maritime. These operational
environments support Line of Sight (LOS) communication.
Palm OS:  Palm OS is a compact operating system initially developed by
the U.S. Robotics’ owned Palm Computing, Inc. for personal digital
assistants (PDAs) in 1996. Palm OS is designed for ease of use with a
touchscreen-based graphical user interface. It is provided with a suite
of basic applications for personal information management.
Personalization:  Services in mobile environment can be catered according
to user’s profile. This is required to let the users easily avail information
with hand-held devices.
228  Glossary

PODS—A Remote Ecological Micro-Sensor Network:  PODS was a

research project in the University of Hawaii that built a wireless network
of environmental sensors to investigate why endangered species of
plants grow in one area but not in neighbouring areas. They deployed
secret sensor nodes, called Pods, in the Hawaii Volcanos National
Park. The Pods consisted of a computer, a radio transceiver and
environmental sensors sometimes including a high resolution digital
camera, relaying sensor data via the wireless link back to the Internet.
Point Coordination Function (PCF):  It is a centralized function used only
on infrastructure configurations.
Power-Efficient Gathering in Sensor Information Systems (PEGASIS):  In
PEGASIS, each node communicates only with a close neighbour. When
a node on the chain receives data from a neighbour, it aggregates the
data with its own data and sends the data to the next neighbour on
the chain. Rather than multiple cluster-heads sending data to the base
station as LEACH, only one node on the chain is selected to transmit
to the base station.
Presentation Tier:  The topmost level of a mobile computing application
is the user interface. The main function of the interface is to facilitate
the users to issue requests and to present the results to the them
meaningfully.
Profile Layer:  The profile layer defines the minimum set of application
programming interfaces (APIs) available on a particular family of
devices. Profiles are implemented upon a particular configuration.
Applications are written for a particular profile and are thus portable
to any device that supports that profile. A device can support multiple
profiles. This is the layer that is most visible to users and application
providers.
Query-based Routing:  In these protocols, the destination nodes query
data through the network. The node(s) containing this required data
send it back to the node that has initiated the query along that path.
RARP (Reverse Address Resolution Protocol):  The RARP protocol is
used by IP to find the IP address based on the physical (MAC address)
address of a computer.
Receiver:  The receiver receives the modulated signals and reverses
the functions of the transmitter component and thus recovers the
transmitted base band signal. The antenna of the receiver is usually
capable of receiving the electromagnetic waves radiated from many
sources over a relatively broad frequency range.
Registering the Care-of-Address:  If a mobile node obtains a care-of-
address from the distant network (foreign), then it should be registered
with the home agent. The mobile node sends a request for registration
to its home agent along with the care-of-address information whenever
the home agent receives the registration request information.
 Glossary   229

RFID:  Radio Frequency Identification (RFID) tag is attached to a product,

an animal, or a person for the purpose of identification and tracking
using radio waves. Some tags can be read from several metres away
and beyond the line of sight of the reader.
Ring Architecture:  A Multi Station Access Unit (MSAU) is a hub or
concentrator that connects a group of computers (“nodes” in network
terminology) to the ring. The nodes are placed along the ring. The
nodes transmit in turn.
Route Optimization:  In the mobile IP protocol, all the data packets to the
mobile node go through the home agent. Because of this, there will
be heavy traffic between the HA and the CN in the network, causing
latency to increase. Therefore, route optimization needs to be carried
out to overcome this problem.
Routing Loop:  By sending the tampered routing packets, an attacker
can create a routing loop. This will result in data packets being sent
around endlessly, consuming both bandwidth and power for a number
of nodes. The packets will not reach their intended recipient and thus
can be considered a sort of denial-of-service attack.
Rumor Routing:  In rumor routing, each node maintains a neighbour list
and an event table. When a node witnesses an event, it adds the event
to its event table. Nodes that have recently observed an event generate
an agent with a certain probability. An agent is a long-lived packet,
roaming through the network and propagating information. Each agent
carries a list of events it has encountered, along with the number of
hops to that event. When it arrives at a node, it synchronizes its list
with the node’s list. The agent travels the network for some number
of hops, and then dies.
Sequential Assignment Routing (SAR):  Sequential Assignment Routing
(SAR) is the first protocol for sensor networks that includes the notion
of QoS in its routing decisions. This protocol creates trees rooted at
one-hop neighbours of the sink by taking QoS metric, energy resource
on each path and priority level of each packet into consideration. By
using created trees, multiple paths from sink to sensors are formed.
One of these paths is selected according to the energy resources and
QoS on the path.
Smart Sensors and Integrated Microsystems (SSIM):  SIMM is a
biomedical project on the artificial retina. In this, a retina prosthesis
chip (prosthesis is an artificial replacement of a missing body part)
consisting of 100 micro sensors is built and implanted within the
human eye. This allows patients with no vision or limited vision to
see at an acceptable level. The wireless communication is required to
suit the need for feedback control, image identification and validation.
SMTP (Simple Mail Transfer Protocol):  The SMTP protocol is used for
sending e-mails.
230  Glossary

SNMP (Simple Network Management Protocol):  The SNMP protocol

is used for the administration of computer networks. The network
manager uses tools based on this protocol to monitor and assess the
network performance.
Snooping TCP:  Snooping TCP protocol improves the TCP performance
by modifying software at a base station while preserving the end-to-
end TCP semantics.
Structure of Mobile Computing Environment:  The structure of a mobile
computing environment indicates a logical division of the underlying
mechanism of the supported functionalities. A simple three-tier structure
for mobile computing environment comprises the Presentation tier, the
Application tier and the Data tier.
Symbian OS:  Symbian OS is a real-time, multitasking, pre-emptive,
32-bit operating system running on ARM processors. Symbian OS from
Symbian Ltd. is designed for mobile devices, with associated libraries,
user interface frameworks and reference implementations of common
tools and runs exclusively on ARM processors.
TCP (Transmission Control Protocol):  TCP is a standard transport layer
protocol. In mobile computing, TCP is possibly the most popular
transport layer protocol. UDP is connectionless and does not guarantee
reliable data delivery. But, TCP on the other hand, guarantees reliable
data delivery between two applications. TCP needs some special
adaptations for use in mobile applications.
TCP (Transmission Control Protocol):  TCP is responsible for breaking
data down into IP packets before they are sent, and for assembling the
packets when they arrive.
TCP Congestion Avoidance:  The congestion avoidance algorithm starts
where the slow-start stops. From this point, the TCP increases its
transmission rate linearly, by adding one additional packet to its
window at each transmission time.
TCP Connection-oriented Service:  TCP is a connection-oriented
protocol. An application requests a “connection” to destination before
transferring data. When two processes (A and B) at different sites want
to communicate, TCP of A informs the TCP of B and gets approval
status from TCP of B. TCP of A and TCP of B exchange data streams
in both directions. State information (the state of transmission) is
maintained at both ends such as sequence numbers, window size, etc.
TCP Fast-Retransmission:  Fast retransmission is based on the observation
that TCP encounters unacceptably long pauses in communication
during handoffs which cause increased delays and packet losses. This
approach addresses the issue of TCP performance when communication
resumes after a handoff.
 Glossary   231

TCP in Mobile Networks:  TCP has become the de facto transport protocol
in the Internet. It has been remarkably successful in supporting many
applications such as web access, file transfer and email.
TCP Port Address:  The client program uses a temporary port number
and the server program uses a permanent (well-known) port number.
These port numbers are used for the identification of the application.
TCP Reliable Service:  TCP is a reliable transport protocol. The implication
of this is that TCP guarantees that data will be delivered without loss,
duplication or transmission errors. For this, TCP uses the following
schemes:  positive acknowledgement, checksum on both header data,
checking sequence numbers to determine the missing segments and
detection of duplicate segments.
TCP Slow-Start:  The slow-start mechanism is used when a TCP session is
started. Instead of starting transmission at a fixed transmission window
size, the transmission is started at the lowest window size and then
doubled after each successful transmission.
TCP Stream Delivery Service:  TCP is a stream-oriented protocol. A
stream is essentially any ordered sequence of bytes so that a group
of bytes (message) sent in one transmission operation (record) is read
exactly as that group at the receiver application.
TCP/IP Application Layer:  This layer provides communication between
applications running on separate hosts. This layer includes protocols
such as http, ftp, telnet, etc.
TCP/IP Data Link Layer:  The functions of this layer are related to the
logical interfacing between a sub-network and an end system. Its
functionalities include encoding schemes and also determining the rate
of signalling determined by the physical layer. It also provides error
detection and packet framing functionalities. It delivers data packets
by making use of lower layer protocols. Ethernet is possibly the most
common data link layer protocol.
TCP/IP Internet Layer:  It mainly performs data routing and ensures
packet delivery at the destination host. It supports communication
between two hosts. Remember that there can be several applications
or processes running on a host. Once a message reaches a host, it
is demultiplexed using the port number at the transport layer for
delivery to the appropriate application. In a nutshell, this layer manages
addressing of packets and delivery of packets between networks using
the IP address.
TCP/IP Transport Layer:  It provides reliable end-to-end data transfer
services. The endpoints of a communication link are the processes.
Thus, to identify the endpoint, not only the computer needs to be
identified, but also the port must be identified since the same computer
may run several applications. The transport layer provides services
232  Glossary

by making use of the services of its lower layer. The details of the
underlying networks are hidden from the application layer by the
transport layer. Sometimes this layer is also referred to as host-to-host
layer. This layer manages the transfer of data by using connection-
oriented (TCP) and connectionless (UDP) transport protocols.
TCP/IP:  TCP/IP is a large collection of a number of protocols based
upon the two important protocols TCP and IP. The Transmission
Control Protocol (TCP) and Internet Protocol (IP) are the two important
protocols of the TCP/IP protocol suite. The TCP/IP protocol suite was
developed by DARPA in 1969.
TCP-F (TCP-Feedback):  The TCP-F (TCP-Feedback) protocol proposes
that the sender can distinguish between route failure and network
congestion. Similar to Freeze-TCP and M-TCP, the sender is forced
to stop transmission without reducing the window size upon route
failure.
TELNET:  It enables the remote log-on facility, by which the user can log-
on the systems directly. Both FTP and TELNET pump in data to the
TCP layer as it is. TCP forwards the same data on the network through
the IP and link layers. As a result, it becomes easy to sniff (secretly
hear) the data by using the publicly available TCP sniffer programs
such as “TCPdump”. Due to this, at present most users use sftp and
ssh which essentially serve to encrypt (and decrypt) data before passing
on to the TCP layer.
Time Division Multiple Access (TDMA):  TDMA is an access method
in which multiple users or sources are allotted different time slots to
access the same physical channel. The available time slot is divided
among multiple sources.
Transaction Model:  It defines the framework for the definition and
execution of transactions. The integrity constraints on a database
are given by well-known ACID properties (Atomicity, Consistency,
Isolation, Durability), meaning that a transaction is an atomic, consistent
and recoverable unit that does not interfere with other transactions that
are executed concurrently. In some application scenarios, such as the
mobile environment, ACID properties turn out to be too stringent for
realizing any meaningful database in practice and therefore require to
be relaxed.
Transaction:  It is a unit of operation that changes a database from a
consistent state into another consistent state. A database is a consistent
state if it satisfies all the defined semantic integrity constraints. A
transaction consists of many read and write operations on the database.
Transmitter:  The input to a wireless transmitter may be voice, video,
data or other types of signals to be transmitted to one or more distant
receivers. This signal is called the base band signal. The basic function
 Glossary   233

of the transmitter is to modulate, or encode, the base band signal onto

a high frequency carrier signal.
Tree-based Protocol:  Tree-based schemes establish a single path between
any two nodes in the multicast group. These schemes require a mini-
mum number of copies per packet to be sent along the branches of
the tree.
Tunnelling and Encapsulation:  Tunnelling establishes a virtual pipe
for the packets available between a tunnel entry and the endpoint.
Tunnelling is the process of sending a packet via a tunnel and it is
achieved by a mechanism called encapsulation. Encapsulation refers to
arranging a packet header and data in the data part of the new packet.
On the other hand, disassembling the data part of another packet is
called decapsulation. Whenever a packet is sent from a higher protocol
layer to a lower protocol layer, the operations of encapsulation and
decapsulation take place.
Tunnelling to the Care-of-Address:  Tunnelling takes place to forward an
IP datagram from the home agent to a care-of-address.
Ubiquity:  The dictionary meaning of ubiquity is “present everywhere”.
Mobile computing allows a user perform computations from anywhere
and at anytime.
UMT Core Network:  The core network is the equivalent of the GSM
Network Switching Subsystem (NSS).
UMT Radio Network Subsystem (RNS):  The RNS is the equivalent of
the Base Station Subsystem (BSS) in GSM. It provides and manages the
wireless interface for the overall network.
UMT User Equipment (UE):  The UE is the name by which a cell phone
is referred to. The new name was chosen because of the considerably
greater functionality that the UE has compared to a cell phone. It can
be thought of as anything between a mobile phone used for talking
and a data terminal attached to a computer with no voice capability.
Unicast:  In this, the message is sent to a single destination node.
Universal Mobile Telephone Standard (UMTS):  UMTS is enlisted as
the third generation technology (3G) with the voice data and non-
voice data (exchanging email, instant messaging and video telephony)
capabilities.
Vehicular Ad Hoc Network (VANET):  It uses the moving automobiles
as the nodes in a ad hoc network.
Virtual Home Environment (VHE):  A user roaming from his network to
other UMTS operators, will experience a consistent set of services, thus
“feeling” to be on his home network, independent of the location. In
contrast, in a 2G network, a user is registered to a visitor location and
is also charged a roaming overhead.
234  Glossary

WAP 2.0:  WAP 2.0 protocol focuses on the schemes combining wireless
applications with the Internet. The WAP 2.0 provides support for the
protocols such as IP, TCP and HTTP. It is also flexible with air interface
technologies and their corresponding bearers.
Windows CE:  Windows CE is a 32-bit multitasking operating system.
Usually, this operating system is delivered on a Read Only Memory.
Wireless Application Environment (WAE):  The application layer
includes the micro-browser on the device, WML (the Wireless Markup
Language), WMLS (a client-side scripting language), telephony service,
and a set of formats for the commonly used data (such as images,
phone books, and calendars).
Wireless LAN Cards:  End-users access the WLAN through WLAN
adapters (wireless network interface cards) in their hand-held devices.
Nowadays, this card is in-built into the motherboards.
Wireless Local Area Networks (WLANs):  WLANs provide connectivity
between computers over short distances via a wireless medium. Typical
indoor applications of WLANs may be in educational institutes, office
buildings and factories where coverage distances are usually less than
a few hundred feet.
Wireless Sensor Networks:  Wireless sensor networks (WSNs) are network
systems containing sensor nodes. The sensor nodes can sense certain
physical characteristics and can be used to capture environmental
information such as temperature, sound, motion, etc.
Wireless Session Protocol (WSP):  The session layer provides HTTP
functionality, with basic session state management, and a facility for
reliable and unreliable data push and pull.
Wireless Transaction Protocol (WTP):  The transaction layer provides
transport services (one-way and two-way) and the related technologies.
Wireless Transport Layer Security (WTLS):  The security layer provides
data security and privacy, authentication, as well as protection against
denial-of-service attacks.
Zone Routing Protocol:  In ZRP, a routing zone comprises a few mobile
ad hoc nodes within a few hops from the central node. Within this
zone, a table-driven routing protocol is used. This implies that route
updates are performed for nodes within the zone. Each node, therefore,
has a route to all other nodes within the zone. If the destination
node resides outside the source zone, ZRP employs a route discovery
procedure.
 Index

Access point, 16 Care-of-address (COA), 66

Acoustic neuroma, 44 colocated, 66
Ad hoc network, 25 discovering the care-of-address, 71
Ad hoc On-Demand Distance Vector registering the care-of-address, 71
(AoDV) protocol, 135 tunnelling the care-of-address, 71
Agent discovery Carrier Sense Multiple Access with
agent advertisement, 67 Collision Detection (CSMA/CD), 7
agent solicitation, 67 Cellular communication technology, 2
Android architecture of the system, 11
advantages of, 197 cell structure, 30
application development, 195, 196 functioning of the system, 3
SDK environment, 194 generations of, 31
software stack structure, 197 2.5 generation (2.5G), 34
Application development, 185 first generation (1G), 31
Application layer, 83 fourth generation (4G), 35
Application tier, 28, 29 second generation (2G), 33
ARP (Address Resolution Protocol), 81 third generation (3G), 34
Atomicity, 102, 103 Centralized environment, 103
Atomicity relaxation, 108 Client-server environment, 104
Code Division Multiple Access (CDMA),
34, 50, 52
Blackberry, 180 Computer networks
Black hole, 142 access arbitration policy, 6
Bluetooth baseband, 21 controller area networks, 4
Bluetooth Link Manager Protocol (LMP), internetworks, 5
21 local area networks, 5
Bluetooth Radio Frequency Communi- bus architecture, 7
cation (RFCOMM), 21 ring architecture, 8
Bluetooth technology, 19, 26 transmission control policy, 6
piconet, 20 Congestion avoidance, 91
protocol stack, 20 Consistency, 102, 103
scatternet, 20 Consistency relaxation, 108
Bridge network, 16, 26 Correspondent Node (CN), 66

235
236  Index

Data dissemination, 107 Equipment Identity Register (EIR),

Data replication, 109 40
Database recovery, 113 Home Location Register (HLR), 39
Datagram, 85 Mobile Switching Centre (MSC),
Denial-of-Service (DoS), 141 39
Destination-Sequenced Distance-Vector Networking and Switching
Routing Protocol, 31 Subsystem (NSS), 39
Directed Diffusion (DD), 159 Operation subsystem (OSS), 40
Distance vector protocols, 127 Radio subsystem (RSS), 38
Distributed environment, 104 Visitor Location Register (VLR), 39
Domain Name Server (DNS), 82 GSM security, 40
Dropping routing traffic, 143
Durability, 102, 103
Durability relaxation, 109 HDML, 187
Dynamic Host Configuration Protocol Hierarchical routing, 158
(DHCP), 75 Home Address (HA), 66
Dynamic Source Routing (DSR) protocol, Home network, 66
132 Hyper Text Transfer Protocol (HTTP), 81
Dynamic topologies, 119

ICMP (Internet Control Message Protocol),

Encapsulation, 67 81
IGMP (Internet Group Management
Protocol), 82
Foreign agent, 66 Indirect TCP (I-TCP), 92
Foreign network, 66 International Mobile Telecommuni-
Freeze-TCP, 95 cations-200 (IMT-2000), 34
Frequency Division Multiple Access Internet layer, 83
(FDMA), 50, 51 Internet Protocol (IP), 6
FTP (File Transfer Protocol), 86 IP addresses, 82
ISO/OSI model, 87
Isolation, 102, 103
General Packet Radio Service (GPRS), Isolation relaxation, 109
34, 41
architecture, 41
services, 41 J2ME, 91
Geographic Adaptive Fidelity (GAF), architecture, 103
160–161 connected device configuration for
Geographic and Energy Aware Routing plug-in devices, 193
(GEAR), 160 connected limited device configu-
Global System for Mobile Communi- ration for hand-held devices, 193
cations (GSM), 35
bearer services, 36
supplementary services, 36, 37 Link State Protocols (LSP), 125
system architecture, 37 Location-based routing, 158
Authentication Centre (AuC), 40 Long Term Evolution (LTE) standard,
Base Station Controller, 39 35
Base Station Subsystem (BSS), 39 Low Energy Adaptive Clustering
Base Transceiver Station (BTS), 39 Hierarchy (LEACH), 160
 Index   237

MAC (Media Access Control), 7, 9, 47, 50 Multiple Access Collision Avoidance

design of, 47 (MACA), 56
protocols, 48, 50
for ad hoc networks, 58
fixed assignment, 50–54 Network access layer, 83
random assignment, 50, 54, 55
reservation-based, 50, 55–57
MAC standard, 57 Operating systems
MANET hybrid routing protocols, 130 basic concepts of mobile O/S, 167
MANET multicast routing protocols, 136 compliance with open standards, 172
MANET reactive protocols (on-demand), library support, 172
130 special constraints of mobile OS, 169
MIME (Multipurpose Internet Mail support for specific communication
Extensions), 81 requirements, 171
Mobile ad hoc networks (MANETs), 116 support for a variety of input
applications of, 120 mechanisms, 170
attacks on, 140 Operational environment, 123
characteristics of, 118
design issues, 122
operational constraints, 120 Packet delivery, 68
schematic model, 118 Palm OS, 174
security issues, 138 Power Efficient Gathering in Sensor Infor-
Mobile commerce, 200 mation Systems (PEGASIS), 160
advantages and disadvantages, 205
features required of a mobile device, 204
Mobile computations, 1 Query-based routing, 57
Mobile computing, 1, 24 Query optimization, 111
applications, 27
characteristics of, 27
local awareness, 27 RARP (Reverse Address Resolution
the three tiers of, 28 Protocol), 81
ubiquity, 27 RFID (Radio Frequency Identification),
Mobile database system 202
ACID properties, 102 Rollback process, 110
applications, 102 Routers, 82
specific requirements, 101 Routing protocols
Mobile environment, 104 in MANETS, 126
Mobile Internet Protocol (Mobile IP), 64 traditional, 124
evolution of, 65 Rumor routing, 159
features of, 70
key mechanisms, 71
overview of, 68 Sensor node, 154
route optimization, 73 bandwidth, 151
terminologies of, 66 density, 151
Mobile Node (MN), 66 identity, 151
Mobile payment systems, 207 movement, 151
Mobile TCP (M-TCP), 94 reliability, 151
Mobile transactions, 110 transmission range, 151
Mobile WiMax standard, 35 Sequential assignment routing, 160
Mote, 150 SMTP (Simple Mail Transfer Protocol),
Multipath-based routing, 157 81
238  Index

SNMP (Simple Network Management Windows Mobile OS (Windows CE),

Protocol), 81 172
Snooping TCP (S-TCP), 94 Wireless Application Environment
Symbian OS, 175 (WAE), 190
Wireless communication system
amplifiers, 10
TCP congestion avoidance, 91 antenna, 10
TCP fast re-transmission, 93 filters, 10
TCP-Feedback (TCP-F) 96 mixers, 10
TCP/IP receiver, 9
application layer protocols, 86 transmitter, 9
architecture, 82 Wireless Datagram Protocol (WDP), 91
in mobile networks, 92 Wireless MAC protocols
operation of, 84 exposed terminal problem, 49
overview of, 79 hidden terminal problem, 48, 49
terminologies of, 80 Wireless networking, 25
TCP window, 87 IEEE 802.11 standards, 12
Telnet, 87 Wireless networks
Time Division Multiple Access (TDMA), fixed infrastructure, 25, 26
34, 50, 51 history of, 15
Transport layer, 83 TCP in multi-hop, 96
Tree-based protocol, 136 TCP in single-hop, 92
Two-phase commit protocol, 110 Wireless sensor networks
applications, 152
characteristics of, 156
Unicast routing protocols in MANETs, design of, 155
121 differences from MANETs, 150
Universal Mobile Telecommunication routing techniques, 158
System (UMTS) 35, 42 schematic model, 149
network architecture, 43 target coverage, 161
core network, 43 types of protocols, 157
Radio Network Subsystem (RNS), Wireless Session Protocol (WSP), 190
43 Wireless Transaction Protocol (WTP),
User Equipment (UE), 43 190
networks, 43 Wireless Transport Layer Security
(WTLS), 191
WLANS, 15
Vehicular Ad Hoc Networks (VANETs), advantages, 18
137 applications, 17
Voice over Internet Protocol (VoIP), 4 architecture, 16

WAP 2.0, 189 Zone Routing Protocol (ZRP), 135

 Fundamentals of Mobile Computing
Prasant Kumar Pattnaik • Rajib Mall
This textbook addresses the main topics associated with mobile computing and wireless networking at a level that
enables the students to develop a fundamental understanding of the technical issues involved in this new and fast
emerging discipline.
The book first examines the basics of wireless technologies and computer communications that form the essential
infrastructure required for building knowledge in the area of mobile computations involving the study of invocation
mechanisms at the client end, the underlying wireless communication, and the corresponding server-side technologies.
The book includes coverage of development of mobile cellular systems, protocol design for mobile networks, special
issues involved in the mobility management of cellular system users, realization and applications of mobile ad hoc
networks (MANETs), design and operation of sensor networks, special constraints and requirements of mobile
operating systems, and development of mobile computing applications.
Finally, an example application of the mobile computing infrastructure to M-commerce is described in the concluding
chapter of the book.
This book is suitable as an introductory text for a one-semester course in mobile computing for the undergraduate
students of Computer Science and Engineering, Information Technology, Electronics and Communication Engineering,
Master of Computer Applications (MCA), and the undergraduate and postgraduate science courses in computer science
and Information Technology.

KEY FEATURES
◆ Provides unified coverage of mobile computing and communication aspects
◆ Discusses the mobile application development, mobile operating systems and mobile databases as part of the material
devoted to mobile computing
◆ Incorporates a survey of mobile operating systems and the latest developments such as the Android operating system

THE AUTHORS
PRASANT KUMAR PATTNAIK, PhD (Computer Science), is Associate Professor at the School of Computer
Engineering, KIIT University, Bhubaneswar. He has more than a decade of teaching and research experience. He is
a Senior Member of the International Association of Computer Science and Information Technology (IACSIT), Singapore.
His areas of interest include mobile computing and cloud computing.
RAJIB MALL, PhD, is Professor in the Department of Computer Science and Engineering at the Indian Institute of
Technology Kharagpur. He has vast practical experience in developing industry-oriented software products. Having an
academic experience of a decade and half in IIT Kharagpur, Professor Rajib Mall has guided several doctoral dissertations
and published over a hundred research articles.

You may also be interested in

Wireless Communications, P. Muthu Chidambara Nathan
Bluetooth Technology and Its Applications with Java and J2ME,
C.S.R. Prabhu and A. Prathap Reddi
Wireless and Mobile Communication, T.G. Palanivelu and R. Nakkeeran

ISBN:978-81-203-4632-1

www.phindia.com 9 788120 346321