BRIEF OVERVIEW OF HOW WIRELESS SENSOR

NODE IS ORGANISED
Sensing sub-system

This helps to integrate one or more sensors

Actuation sub-system

Enforces the sensor to act on the environment

Processing sub-system

Responsible for data processing, creating and transmitting

the network packets, data aggregation etc
 Power management sub-system

 Ensures that all the components are supplied with the necessary
power

 Communication sub-system

 Responsible for creating data streams and change the data streams
to packet for transmission over the network

 Security sub-system

 Enforces the security of the data coming from one sensor node to
another so that the said data is not compromised
 Wireless sensor network
 The internet interfaces human beings with the virtual world

 Almost unlimited internet access

 Allows human to control his or her environment

 It embeds a number of sensor nodes with the aim of interacting with

the environment

 The devices are internetworked so that they can cooperate to perform

high level sensing tasks

 Nodes collaborate with each other to support in-network processing,

which is away of significantly reducing the amount of network data
ILLUSTRATION OF NODES ON A NETWORK
 APPLICATIONS
Wireless sensor network has a large range of where its applicable. Below are some of the
areas;

 Structural health monitoring : inspection of big buildings to ascertain the integrity hence
avoiding damages

 Traffic control: monitoring urban congestions by predicting traffic flows etc

 Health care: mainly applied in Tele Medicine where medicine is packaged and
delivered

 Pipeline monitoring: detect and rectify damages in pipes

 Precision agriculture: manage resources like pesticides, water etc in large scale farm

 Underground mining: safety of miners and mines is key for monitoring such as air quality
etc.
Node Architecture
The architecture deals with how each
component of the wireless sensor node should
be designed.

Each design choosing the hardware,

connecting them, how the signal should flow,
the interconnection of the connections ect
 Sensor selection and integration
 The quality of the data we extract from the physical environment depends on the quality
of the sensors, how the sensors are integrated into the holes of the sensor nodes.

 Accuracy

 Sensitivity

 Reproducibility

 Span

 Resolution

 Selectivity

 Response time

 Self-heating
 Operating system
Every wireless sensor node is built with an Operating system. We
need to understand the;

Functional properties

Non-Functional properties

Proto Type
Tiny OS

Contiki
 Physical layer (point to point connection)
 Responsible for establishing a direct link between two sensor nodes or point to
point.

 The three main building blocks are;

 Transmitter: generates data to be communicated

 Channel: may affect the signal in different ways like noise, attenuation

 Receiver: Detect signal, isolate the signal, amplify signal demodulate the signal
etc

 Since we are dealing with digital signals, we may now talk about other block
such as source encoding, channel encoding, modulation, demodulation and
signal propagation
 MAC (Medium Access Control) protocol
It ensures that all nodes share the medium in a fair;

avoiding collision of packets

This is done through;

Schedule base

Contention base

hybrid
 Network Layer
 If two nodes are unable to communicate for some reason, then there is
need to set up a third node to allow communication to proceed.

 If two nodes would like to communicate but for some reason the
distance between them is too far from each other and their
communication ends does not allow them to communicate, then we
need to add an intermediate node which will receive packets and
forwards the packet to the desired node.

 To reduce the energy consumption during communication between

nodes on a network, there is need to reduce on the distance between
nodes
Here we deal mainly with two kind of
topologies
Flat Topology:
All nodes are equal, play the same roles,
no hierarchy.
Routes are discovered on demand. Its
hopes are highly elastic.
Not efficient for long term deployment
 Hierarchical topology:

 Straight routes of communication between nodes.

 Communication is from Parent to child and vice versa.

 Routes are set before nodes are placed on the network.

 Each node knows who their parents are and they transmit data only to
this parent.

 The cost of communication in this kind of network is high at start but

once communication routes have been established, its faster and less
costly.
Hierarchical topology
 Applications (Structural health monitoring)
This deals with monitoring of integrity of complex
structures such as buildings, planes, bridges etc

Once a bridge collapses, there is usually an

argument between the builders and the
government or people whom the building belongs
to
Structural health monitoring guidelines used
by Civil engineers
 Traditionally: visual inspection done everyday. This becomes
labour intensive, tedious, inconsistent and subjective

 Basic inspection: done at least once a year by experts

 Detailed Inspection: At least every five years on selected

bridges

 Special Inspection: according to technical needs. May only

concentrate on a certain section of the structure but not a
whole of it
Applications (Structural health monitoring)
 Classification of inspection
Local Inspection techniques

Detecting small fractures in the structure

Global Inspection techniques

Looks at damages that are large enough to

affect the entire structure
 Golden bridge gate
This is an example of the network sensor
deployment
 64 wireless sensor nodes deployed on this bridge

 The network monitors ambient vibrations synchronously

Goal

 Determine the response of the structure to both ambient and extreme condition

 Compare actual performance to design predictions

 Measure ambient structural accelerations to the wind load

 Measure strong shaking from a potential earthquakes

 The installation and the monitoring was conducted without the disruption of the
bridge’s operation
 Applications (Traffic Control)
 Ground transportation is a vital and complex socio economic
infrastructure

 It provides support to the supply chain, emergency response and public

health

 All these are usually affected by congestion which results to pollution,

time wastage, organisation loose money due to late arrivals of staff

 Building new road may not be the best solution.

 There is need to put in technology like sensors to predict traffic, to

provide alternative routes etc
 How can sensing help with congestion

Use the sensors to know the number of cars which are on the
street

For example;

A moving car can disturb the distribution of the magnetic field

by producing its own magnetic field or by cutting across it

The magnitude and direction of disturbance depends on

speed, size and density of the car
 Applications (Tele medicine/Gastroparesis)
 When you fall sick, you may need to go see a doctor.

 This need time off work etc

 Wireless sensor has made this much easier where sensors are implanted
onto the body of the patient and the patient does not have to move to
the doctor.

 The condition of the patient is transmitted remotely to the doctor for

analysis and interpretation.

 Treatment medicine is then dispatched to the patient through sensors

such as drones
Applications (Wireless Motility Sensor)

 The sensor is imbedded into the receiver to detect temperature etc

 It helps with monitoring pressure

 Applications (precision agriculture)

Monitors the quality of the agricultural products,

climate change, weed and pests
Also helps to determine how much fertilizer is
required, how much water resources and where the
water is needed
It also helps to measure the quality of the soil and
the moisture of the soil
 Applications (Active Volcano)
The earth Volcanoes are hidden from our views

To monitor this, we need expensive devices that

are difficult to move

The deployment of these devices require vehicles

and helicopters

Data storage must be retrieved on a periodic

basis
 WIRELESS SENSOR NODE ARCHITECTURE

Design decisions:
The existence of an object can be explained by four
different causes;
Material cause: this could be the input material

Formal Cause: this could be the design

Efficient cause: this could be the processing

Final Cause: this could be a goal

Node Architecture
The sensing subsystems
 Analog to Digital Converter
Usually, the first decision to make is what type of
sensor is needed

However the second decision is the conversion of

analog to digital

ADC converts the output of a sensor which is

continuous and analog signal to digital signal
 Suppose we are given the analog signal which is continuous on both
time and amplitude axis

 The analog signals have to be transformed into bits

 The analog signals will have to be sampled at regular intervals. The

interval is determined by the frequency of the signal.
 There is need to map the amplitude of the analog signal to the nearest intersection
point on the graph

 In case the two points are hard to choose, then we introduce an error in the
mapping which we call it quantization.

 Quantization Q is computed as;

 Epp is the peak to peak analog voltage

 2M M is a number of bits assigned to each symbol (ADC’s resolution in bits)

 The processor sub-system
 It connects all the other sub-systems and other additional peripherals

 The main function is to execute instructions pertaining sensing, communication and self
organising

 Some of its components are; processor chip, non-volatile memory, an active memory, internal
clock

THREE TYPES OF PROCESS ARCHITECTURES

1. Von Neumann architecture

2. Harvard architecture

3. Super-Harvard architecture
 Von Neumann architecture

 This is the simplest architecture

 It’s the most compacted architecture

 It has two building blocks (process subsystem and the Memory

subsystem)

 The memory subsystem and the process subsystem are always

interconnected in parallel with two buses

 One unidirectional bus is to address the data in memory and the other
bidirectional bus is for exchanging data
Provides a single memory space for storing program
instructions and data

Provides a single bus to transfer data between the

processor and memory

Slow processing speed because each data transfer

requires a separate clock
 Harvard architecture
This is the second generation architecture

Two separate memory are used to separate data

and instructions

The cost involved is almost double the one in Von

Neumann architecture

The size is larger compared to that in Von Neumann

architecture

This architecture is faster

 Super-Harvard architecture

Tow additional components are introduced to this

architecture

Cache: which stores frequently accessed data

I/O controller: to support direct memory

This is much more faster

The cost incurred is very high

The program memory can also be used to store data

PROCESSORS
 Microcontroller
This this the first type of processor architecture

Generally we don’t use microprocessors for wireless

sensor network because they are very expensive

It has its own CPU, RAM, ROM, I/O interfaces, clock
generator, one or more internal analog to digital
converters, serial communications interfaces.
Advantages
 It is suitable for building computationally less intensive, stand alone
applications because of its small size, low-power consumption and low cost

 Easy to program and debug because of higher-level programming languages

Disadvantages
 It is not as powerful and as efficient as some custom made processors

 Some applications may prefer to use architecturally simple but cost effective
processors
 Digital Signal Processor
 It processes discrete signals with digital filters

 The filters helps to minimise the effect of noise on a signal

 Analog signal processing requires complex hardware components

while digital signal processing requires simple adders, multipliers and
delay circuits

 Digital signal processing are highly efficient

 Most Digital signal processing are designed with the Harvard

Architecture
Advantages
Powerful and Complex digital filters can be realised with the
commonplace digital signal processing

Useful for applications that requires the deployment of nodes in

harsh physical settings

Disadvantages
Some tasks requires protocols that requires periodic upgrade or
modification
Application-specific integrated circuit
 This is an integrated circuit which can be customised for a specific application

 It makes used of two types of designs;

 Fully customised:

 comes with fully implemented integrated circuits. This means it comes with little flexibility and its
optimised for efficiency.

 If you wanted to change the program logic may not be possible since it does not support
modification

 half customised:

 used to reconfigure after production.

 During the design, the application specific are intended for a particular node that was
developed.
 Field Programmable Gate Array
To differentiate between Field Programmable Gate array and
Application-specific integrated circuit is that;

Field Programmable Gate array are more complex in design

and more flexible to program

Field Programmable Gate array are programmed electrically

by modifying the packaged part

Programming is done with the help of the circuit diagrams and

hardware description languages
Advantages
 Higher bandwidths

 Flexible in their application

 Supports parallel processing

 Works with floating point representation

 Greater flexibility of control

Disadvantages
 Its complex

 The design and realisation process is costly

COMMUNICATION INTERFACES
 Communication Interfaces
 The decision as to which of the interfaces are suitable for connectivity
is dependent on speed of communication between the different
nodes, the communicating devices and the available space for
communication

 Energy efficient and faster data transfer between the subsystems of the
wireless sensor network is very important. However;

 The size of the node may put restrictions onto the system bus

 Communication is faster on a parallel bus compared to serial

transmission

 A parallel bus may need more space

 Types of serial buses
Serial peripheral interfaces (SPI)

General purpose Input / Output (GPIO)

Secure Data Input / Output (SDIO)

Inter-Integrated Circuit (I2C)

The most commonly used buses in WSN are SPI and I2C
because it supports both full duplex and half duplex
communication