TOPIC 5

NETWORK STANDARDS
Introduction

Standards ensure that the things we use every day work together reliably. As computer networks
have grown in size, importance, and popularity, a wide variety of standards have developed to
make it easier for network engineers to ensure that computers and networks will connect
together and communicate reliably and efficiently

There are two types of standards in the network industry:

• Defacto standards.

• Dejure standards.

De facto standards arise in the industry by virtue of wide spread commercial and
educational use. They are often proprietary and may be unpublished and unavailable to outside
vendors. Unpublished and unavailable standards are known as closed system standards.
Published and accessible standards are known as open standards; examples of de facto
standards such as IBM’s systems network architectures.
De jure standards are non-proprietary; no single company makes them or owns rights to them.
They are developed with the intent of providing for greater connectivity and interoperability by
making specifications public so that independent manufacturers can build to such specifications.
Examples of non-proprietary de jure standards are OSI reference model, TCP/IP protocol suite,
the IEEE 802 standard. Examples of non- proprietary de jure standards are OSI reference
model, TCP/IP protocol suite and the IEEE 802 standards

Open and Closed Standards

The standards that are established by standards bodies are generally open standards, which means
that anyone in the industry can obtain and use the technical specifications for the standards in
order to implement them or design new products and software that use the standards. They
are also non-proprietary, which means they aren't the property of any particular institution.
Developers and network engineers appreciate open standards; one of the reasons for the
exponential growth of the Internet is that it's based entirely on open, n o n -proprietary standards.
No one owns the Internet.
Other standards are proprietary, meaning that they were developed and are owned and controlled
by a particular company. The company makes money by charging fees for letting people use its
standards and build new products that meet its standards. These standards are also called closed

Apple Computer has historically relied on closed standards in its computing products;
if you wanted to use Macintosh computers, you had to get all your equipment and accessories
directly from Apple.
 THE OPEN SYSTEMS INTERCONNECTION
(OSI) MODEL

To better understand how computer network works we will use a network model. This
model will help to understand and identify the various parts that make up the network
and understand how the parts work together. The OSI reference model is not a
protocol in itself, but rather a conceptual architecture with which to build
communication systems around. It does not promote any one vendor’s product or
technology. The model divides the communication process into seven interdependent
layers.

7 Application
6 Presentation
5 Session
4 Transport
3 Network
2 Data Link
1 Physical

Each layer is responsible for a specific function, and when combined with other layers,
enables data communications to take place. The grouping of these programs in the OSI
model is called the protocol stack. It is extremely important to understand what each
layer defines, and how the layers relate to other layers, and to the model as a whole.
The seven layers of the OSI model can be split into two halves, those which provide
interconnection services and those which provide internetworking services. Each layer
within the model provides a set of services to the layer above and enhances the service
provided by the layer below
The Interconnection Layers
Interconnection group of standards makes up the bottom four layers of the OSI model,
which are known as the physical, data link, network and transport layers.

The physical layer defines the functional, procedural and physical interfaces of
communication links between equipment. For example, plug specifications, and pin
allocations.

The data link layer adds error-checking information and formats data for physical
transmission.

The network layer provides routing and multiplexing services.

The transport layer includes error detection and correction as well as multiplexing. Its
basic function is to enhance the quality provided by the network layer below, if this is
necessary.

The Internetworking Layers

The internetworking group includes the top three layers of the OSI model and basically
provides the support services for the user applications. They are known as the session,
presentation, and application layers.

The session layer provides the organization, synchronization and timing of the
exchange of the data between end systems.

The presentation layer is concerned with now the information to be exchanged. This
includes resolving character set differences, such as ASCII to EBCDIC, providing text
compression and encryption/decryption services.

The application layer provides support for the user applications, which wish to exchange
information. (i.e. file transfer

Layered Architecture of OSI

The OSI model has seven layers. Number of layers in any model is derived on following
principles.

 A layer should be created where a different level of abstraction is needed.

 Each layer should perform a well define function.

 The function of each layer should be chosen with an eye towards defining
internationally standardized protocol
 The layer boundaries should be chosen to minimize the information flow across the
interface.
 The number of layers should be large enough that distinct function need not be thrown
together in the same layer out of necessity, and small enough that the architecture does
 not become unwieldy
Advantages of Layered Network Architecture

 Provide modular approach for any network architecture

 A new layer can be introduced any time (if required) without interfering other layers.

 A layer can be removed easily if its functions become obsolete.

 Modification to a particular layer can be done without interfering other layers.

Disadvantages of Layered Network Architecture

 Increases the address overhead in data packet as it travels from bottom layer to the top
layer.

LAYER FUNCTIONS
Physical layer
The data units at this layer are called bits. This layer defines the mechanical and electrical
definition of the network medium and network hardware. This includes how data is merged
onto the cable and retrieved from it. An example of the Physical layer is RS-232.
The main network device found at the Physical layer is a repeater. The purpose of a repeater (as
the name suggests) is simply to receive the digital signal, reform it, and retransmit the signal. It
should be easy to see that this has the effect of increasing the maximum length of a network,
which would have been reached due to signal degradation if a repeater was not available. It
should also be possible to anticipate the requirements of a repeater, it is simply regenerating
cleaner digital signal so doesn't have to understand anything about the information it is
transmitting, processing on the repeater is non-existent
Data link layer
The data link layer groups the bits that we see at Physical layer into Frames. Primarily
responsible for error-free delivery of data we see the DATA LINK layer as split into two, the
Logical Link Control and Media Access Control. Although this split may not be necessary for
the exam it is important to be aware if you see reference to the LLC that this is actually
part of the data link layer
The main network device found at the data link layer is a bridge. This device works at a
higher layer than the repeater and therefore is a more complex device. It has some understanding
of the data it receives and can make a decision based on the frames it receives as to whether it
needs to let the information pass, or can remove the information from the wire. This means
that the amount of traffic on the cable can be reduced and therefore the usable bandwidth
increased.
Network Layer
The unit of data at the network layer is called a packet. This layer is responsible for
ROUTING. It examines network addresses in the data instead of physical addresses seen in
the Data Link layer. It makes a decision as to where to route the packet based on information
and calculations from other routers, or according to static entries in the routing table.
The main device found at the Network layer is a router.

Transport Layer
The transport layer is responsible for the reliable delivery of data. The mechanism for this is
often a connection-oriented delivery where both communicating devices go through a process
where they come to expect data from the other. Once the transfer of data is complete they go
through a 'teardown' phase, which breaks the connection between the machines. There is more
overhead involved in this reliable delivery of data, but it ensures that errors are recognized and
recovered from
Some applications do not require this level of reliability and make use of connectionless
delivery where the above connection-building process is not performed. This in turn reduces the
level of overhead involved. For example, a file transfer over the Internet would probably make
use of reliable connection oriented delivery, whereas real time audio information (e.g. an
Internet radio show) can transmit using connectionless delivery. If some information is missed
the first time, then being realtime audio it is lost forever (you don't want it 5 seconds late and out
of sequence).

Session Layer
This layer is about establishing connections, authenticating access, co- ordination and
synchronization. The session layer is responsible for managing dialog between two devices,
for example the session layer determines whether the communication will be simplex, half-
duplex or duplex.
Presentation Layer
This layer is responsible for ensuring the data is in a format that is understood by both sender and
recipient. This layer is also responsible for data encryption and data compression

Application Layer
The Application layer is probably the most easily misunderstood layer of the model. This layer
deals with providing the network services to the end user, but is not the software running on their
machine (Word forinstance). Server messaging blocks (SMB s) exist at the application layer.
Each layer, with the exception of the physical layer, adds information to the data as it travels
from the application down to the physical layer. This extra information is called a header.
TCP/IP MODEL/ INTERNET MODEL

Has four layers;

 Application-needed to support variety of applications. Separate module

supports each type of application
 Host-to-Host (transport)- Reliable exchange of data between applications
 Internet-Internet layer handles tasks similar to network access layer, but
between networks rather than between nodes on a network. Uses IP for
addressing and routing across networks
 Network Access-Exchange of data between end system and network.
Address of host and destination.
 Physical-comparable to the application, presentation, and session layers of
the OSI model all combined into one. It provides a way for applications to
have access to networked services

NETWORK PROTOCOLS
A Protocol suite (also referred to as protocol stack) is a hierarchical set of protocols that
coordinates to affect a communication process. Protocols are the agreed upon ways in
which computers exchange information. They are sets of basic steps that both parties must
perform in the right order. For example the sending computer must do the following: -
1. Break the data into packets
2. Addressing information to the packets identifying the destination computer
3. Deliver the data to the network card for transmission over the network
The receiving computer must -:
1. Accept the data from the network adapter
2. Remove the transmitting information that was added by the transmitting computer.
3. Reassemble the packets of data into the original message Protocol suite can be
perceived to as a software implementation of the OSI upper layer specification. Various set of
protocol implemented by different vendors exists. This includes TCP/IP,
NETWARE IPX/SPX,NWLINK etc.

TCP/IP PROTOCOL SUITE

As the world of networking evolves, so to the communication standards. The Internet has been
one of the largest and most profound influences in transforming the PC of the 90s; we turn to
the World Wide Web for everything —from banking, to shopping, world via e-mail. With
the use of the Internet comes a renewed interest in the protocol TCP/IP, the language of
the Internet.
TCP/IP is not a single network protocol (remember that a protocol is the set of rules that two or
more computers agree on to facilitate communications), but a suite of protocols and
applications. TCP/IP is extremely well suited for LANs and WANs; it is quite versatile and
a good choice for many configurations. At least for the foreseeable future, TCP/IP will be the
dominant network communications suite.

The original goal behind the TCP/IP communications standard was to create an easy
method to establish connections between many dispersed computers. Back in the mid-1960s,
compute networks only existed in the minds of a few scientists. It was the scientists’ dream
to create a computer communication system that would facilitate collaborative research
across universities worldwide. In the late 1960s, the United States Defense Department’s
Advanced Research Project Agency funded a project to do just this.

When the concept of this connected network was first proposed, standards for computer to
computer communications did not exist. Therefore, it was necessary to develop a new protocol to
handle the diversity of the various computer languages. Because it was impossible to re-
engineer existing systems to a new standard, some compromises had to be made, especially in
relation to security. Also, because the primary long distance links would be voice grade
telephone systems, the protocol had to be as efficient as possible in order to keep
retransmission requirements to a minimum. Due to the many demands for a fast, efficient
low-overhead suite of protocols, Transmission Control Protocol / Internet Protocol, or simply
TCP/IP, was developed. Since its conception, TCP/IP has been improved and enhanced, and
is now considered by many the protocol of choice for today’s networks.

Network Interface Layer

At the base of the model is the network interface layer. This layer puts frames on the wire and
pulls frames off the wire.

Internet Layer Protocols

• IP
Provides connectionless packet delivery for all other protocols in the suite. Does not
guarantee packet arrival or correct packet sequence?
• ARP
Provides IP address mapping to the MAC sub layer address to acquire the physical MAC control
address of the destination. IP broadcasts a special ARP inquiry packet containing the IP address
of the destination system. The system that owns the IP address replies by sending its
physical address to the requester. The MAC sublayer communicates directly with the network
adapter card and is responsible for delivering error-free data between two computers on a
network.
• ICMP
Provides special communication between hosts, allowing them to share status and error
information. Higher-level protocols use this information to recover from transmission
problems. Network administrators use this information to detect network trouble. The ping
utility uses ICMP packets to determine whether a particular IP device on a network is functional.
• ROUTING INFORMATION PROTOCOLS (RIP)
Internet RIP performs route discovery by using a distance vector method, calculating the number
of hops that must be crossed to route a packet via a particular path. Although it works well
in localized networks, RIP presents many weaknesses that limit its utility on wide area
inter- networks.
• OPEN SHORTEST PATH FIRST (OSPF)
The OSPF protocol is a link state route discovery protocol that I designed to overcome the
limitations of RIP. On large inter-networks, OSPF can identify the inter-network topology and
improve performance by implementing load balancing and class of service routing.
• IGMP
Provides multicasting, which is a limited form of broadcasting, to communicate and manage
information between all member devices in a multicast group. IGMP informs neighboring
multicast routers of the host group memberships present on a particular network. Windows
2000 supports multicast capabilities, such as Windows 2000 Server NetShow Services, that
allows developers to create multicast programs.

Transport layer protocols.

Transport layer protocols provide communication sessions between computers. The desired
method of data delivery determines the transport protocol. The two transport layer protocols are
Transmission Control Protocol (TCP) and User Datagram Protocol (UDP
• TCP
Provides connection-oriented, reliable communications for applications that typically transfer
large amounts of data at one time or that require an acknowledgment for data received. TCP
guarantees the delivery of packets, ensures proper sequencing of data, and provides a
checksum feature that validates both the packet header and its data for accuracy.
TCP provides transport functions that ensure the total amount of bytes sent is received
correctly at the other end. TCP provides for reliable host- to-host communication and has
mechanisms built-in to inform you if a transmission fails (for example, if a router is down and
no other routes to the destination exist). Reliable transportation is a good thing. With TCP, you
don’t need to worry about the data getting through; it is like sending a registered letter in that
you are sure that it reaches its intended recipient. However, there is a cost for reliable
transportation;
TCP has a lot of overhead that increases network traffic, and decreases performance of the
network as a whole. It is necessary for many forms of communications and not as necessary for
others. The next section covers the protocol used when reliability is not a primary issue.
TCP is used by a number of other programs and utilities contained within the TCP/IP suite.
Specifically, file transfers using FTP, terminal emulation using Telnet, and e-mail delivery
using SMTP all use TCP as their transportation methods.
Recall from the previous discussion of NetBIOS that its native trans-port protocol is NetBEUI,
which is nonroutable. However, NetBIOS can be combined with TCP/IP, which is routable.
For NetBIOS to function over TCP/IP, NetBIOS must use ports. For example, the NetBIOS
name service that I discussed uses port 137 with the UDP protocol (covered later).
TCP communication between two hosts is much like human communication via the
telephone in that the complexities between callers is not as important as a clear, uninterrupted
connection. One of TCP’s responsibilities is to create an inter-face between the host-to-host
layers and upper layer processes. The IP address takes you to the host, but how do you get to
the proper process running on that host?
To identify specifically where data is to be transmitted, a port number identifies each
protocol. The port number is appended to the IP address of the computer.
For example, if you download a file from the Internet using the FTP protocol, TCP
establishes a connection between your computer and the Internet computer using Port 21
[xxx.xxx.xxx.xxx/port 21]. A large number of port numbers, often referred to as well-
known ports, are reserved for specific protocols. The most commonly used port numbers are:

USED

File Transfer

Simple Mail

Domain Name
Dynamic
 Configuration Protocol
69 UDP TFTP Trivial File Transfer
Protocol
80 TCP HTTP World Wide Web HTTP
110 UCP POP3 Post Office Protocol 3

.
• UDP
Provides connectionless communications and does not guarantee that packets will be delivered.
Applications that use UDP typically transfer small amounts of data at one time. Reliable delivery is
the responsibility of the application.
Like TCP, UDP is a transportation mechanism, but unlike TCP, UDP is unreliable. This is not to
say that UDP is a lower-class mechanism, but simply that the data it transports is not of the nature
that you need confirmation it has been delivered. UDP uses a best effort delivery method, and
does not inform the other protocols if a packet is undeliverable. The primary use for UDP
transmission is speed. Because error checking is omitted, there is very little overhead associated with
the UDP protocol. Among the services that use UDP are Trivial File Transport Protocol (TFTP),
Simple Network Management Protocol (SNMP), Domain Name Services (DNS), and Network File
Systems (NFS). A good example of UDP is live video and audio services on the Internet. If a datagram
is lost during video or audio trans-mission, there simply isn’t enough real time to retransmit the
dropped data and remain live.

Application layer protocols

At the top of the model is the application layer, in which applications gain access to the network.
There are many standard TCP/IP utilities and services in the application layer, such as FTP, Telnet,
Simple Network Management Protocol (SNMP), DNS, and so on.
• FILE TRANSFER PROTOCOL (FTP)
FTP is the standard method of sending files between computers over a TCP/IP-based network.
Most commonly found on the Internet, FTP is reliable and offers in the real world some basic
security features. The FTP protocol provides a command set that permits such things as connecting
to a network using a user name and password, and navigating the directory structure. FTP
can be invoked from a command line environment, a graphical interface, or even from a Web
browser by preceding the server address with ftp:// (For example, you can enter ftp://microsoft.com to
access publicly available programs from Microsoft). Many FTP sites permit anyone to use
the login name anonymous to access and download files. The anonymous FTP directory is isolated
from the rest of the system and generally does not accept uploads from users. Many of these
anonymous sites use your e-mail name as the password so they have some indication of who is
accessing the files.
Unlike e-mail programs that require graphic and program files be attached, FTP is designed
to handle files directly and does not add the overhead of encoding and decoding the data. FTP
services can be installed on most Unix, Windows 95/98/2000, and NetWare systems. As mentioned
above, Trivial File Transfer Protocol, or TFTP, uses UDP as its transportation mechanism and
does not have directory or password capabilities.
• HYPERTEXT TRANSFER PROTOCOL (HTT)
HTTP is the communication protocol that connects servers on the World Wide Web and transmits
Hypertext Markup Language (HTML) pages to a client browser (such as Netscape Navigator or
Internet Explorer). Most browsers assume the ad-dress you enter has HTTP connections; the
browser automatically inserts the http:// portion for you. The HTTP protocol is undergoing many
changes to make it more efficient and useful so there is no need to discuss the protocol in much detail
here.
• POST OFFICE PROTOCOL VERSION 3 (POP3)
Most client e-mail packages in use today use Post Office Protocol 3, or POP3, especially with the
proliferation of dial-up access. POP3 provides a message store that holds all incoming mail on a
central server until the recipient connects to the server and downloads it. POP3 is very
simplistic, and offers little security beyond basic user name and password. When a
connection is established and verified, all pending messages and file attachments are downloaded at
the same time. This way, you can read your e-mail and view any file attachments without
having to re-main connected to the Internet.

• SIMPLE MAIL TRANSPORT PROTOCOL (SMTP)

Before e-mail reaches your central holding point and waits to be retrieved from a POP3 client, the
servers that transport the e-mail data around the Internet need a protocol. Simple Mail Transport
Protocol, or SMTP, is just such a standard. SMTP is a TCP/IP protocol that defines the message
format, and the methods to store and forward the mail. Servers use SMTP to route messages
throughout the Internet to mail servers which then provide message storage for incoming mail.
•SIMPLE NETWORK MANAGEMENT PROTOCOL (SNMP)
With the ever-increasing size and complexity of Network systems, a standardized method of
managing the multitude of components has become a necessity. In 1988, the Internet Engineering
Task Forced developed the Simple Network Management Protocol, or SNMP, to enable simple,
but extensible network management.
Devices enabled with the management client components monitor various network devices and
gather information about the components. These client components are known as management
agents. A PC running special management console software is also required to complete the
SNMP process. The management agents are programmed with certain thresholds, which are
minimum or maximum acceptable values. When one of these values falls outside the range of a
threshold, the agent issues a trap, thus notifying the management console of the problem. SNMP is
based on a GET/SET paradigm. The Management Console can request information from an agent
(GET) or, if sufficient security is given, alter the value of a threshold (SET).

Advantages of TCP/IP
(a) Broad connection among all types of computers and servers.
(b) Direct access to the global Internet.
(c) Strong support for routing.
(d) Includes simple networks management protocol support (SNMP)
(e) It support for (DHCP) to dynamically assign client IP address.
(f) Support for the windows Internet Name service (WINS) to allow name browsing among
Microsoft client server s.
(g) It has support for most other Internet Protocol such as HTTP. (h) It allows networking
between organisation.

Disadvantages of TCP/IP
(a) Difficult to setup
(b) Relatively high overheads to support connecting and routing
(c) Slower speed than IPX and NETBEUI

NWLINK
This is Microsoft’s implementation of Novell’s IPX/SPX protocol stack. Used in Novell
NetWare. Actually IPX is the protocol. IPX is included with each Microsoft Windows NT
to support connection totheNovellNetworkservers. This makes migration between platforms
very easily. It also makes need form a complete cut-over from one networking standard to
another very easily. NWlink does not allow file and print sharing and from NetWare
clients or servers.
Advantages
(a) Ease of setup
(b) Support for routing between networks
(c) Greater speed than for TCP/IP implementation
(d) Ease of connection to installed NetWare server and client s.

Disadvantages
(a) Slower than NetBEUI
(b) Doesn’t support NetWare management protocols
(c) It lacks an effective centralized network addressing scheme to ensure that the network don’t use
the same address numbers as where is with TCP/IP

NetBEUI
NetBEUI stands for NetBIOS Extended user interface. (Net Bios network Basic Input/output
System). NetBEUI was developed for work group of 20 -200 computers. IBM originally developed
NetBEUI for departmental LANS. It is extremely easy to configure requiring very little input from
the administrator. NetBEUI acceptance is limited to Microsoft and IBM personal computer
networks and s u f f e r s from one major problem. NetBEUI is not routable so all
communication using this protocol is limited to the local network. Hence, NetBEUI does not
scale at all. However, NetBEUI is fast. These parameters make this protocol ideal for small LANs.

Advantages
a) High speed on small networks
b) If easy to setup and implement if largely self-tuning (self-tuning features)

Disadvantages
a) It cannot be slotted between networks (unsuitable for large network)
b) Very few management and maintenance t ool s for NetBEUI which makes it difficult to
troubleshoot.
c) It offers very little class -plat form support.

IPX SPX (Internet packet exchange protocol

/sequence packet protocol
IPX SPX is the network protocol used by Novell NetWare clients to connect to NetWare
servers under Windows 98 and 2000. IPX SPX is very easy to install and configure. It can be used
for both large and small network. It can operate between both NetWare and ms network operating
system, thus very attractive choice for multi-vendor networks.