STRU TUR , OOR INATION, S NSING AN ALLO ATION IN OGNITIV ...

STRUCTURE, COORDINATION, SENSING AND

ALLOCATION IN COGNITIVE RADIO VANETS

SHAHID HUSSAIN ABBASSI

A Thesis

Submitted in Partial Fulfillment of the

Requirements for the Degree of

Doctor of Philosophy

DEPARTMENT OF ELECTRICAL ENGINEERING

AIR UNIVERSITY

2016



Ph.D. Dissertation

SUBMITTED BY

SHAHID HUSSAIN ABBASSI REG. NO. Ph.D.-EE-091313

SUPERVISOR

PROF. DR. IJAZ MANSOOR QURESHI

DEPARTMENT OF ELECTRICAL ENGINEERING AIR UNIVERSITY

ISLAMABAD

February, 2016

CERTIFICATE OF APPROVAL

Department of Electrical Engineering

It is hereby certified that Shahid Hussain Abbassi (Reg # Ph.D.-EE-091313) has successfully completed his dissertation.

_____________________________

Dr. Ijaz Mansoor Qureshi

Air University Supervisor

____________________________ ____________________________ Dr. Fida Muhammad Khan Dr. Syed Ahmed Pasha Internal Examiner 1 Internal Examiner 2 Guidance and Evaluation Committee Guidance and Evaluation Committee

____________________________ ____________________________ Dr. Adnan Omer Dr. Abdul Jalil External Examiner External Examiner Guidance and Evaluation Committee Guidance and Evaluation Committee

____________________________ ___________________________

AVM Saleem Tariq Dr. Zafar Ullah Koreshi Chair Department Senior Dean



Ph.D. Dissertation

SHAHID HUSSAIN ABBASSI REG. NO. Ph.D.-EE-091313

SUPERVISOR

PROF. DR. IJAZ MANSOOR QURESHI

FOREIGN RESEARCH EVALUATION EXPERTS

Dr. Amir Hussain, Divisional PhD Director, University of Stirling, Scotland, UK

Dr. Wen-Hsien Fang, Professor and Chairman, Department of Electronic and Computer Engineering, National Taiwan University of Science and

Technology, Taipei, TAIWAN

DEPARTMENT OF ELECTRICAL ENGINEERING

AIR UNIVERSITY ISLAMABAD

2016

ABSTRACT

With the increase in population, vehicle traffic has also increased on roads. This has caused an

increase in accidents, due to which thousands of people lose their lives and millions get injured

annually. Hence, a foolproof and secure Vehicular Ad-hoc Network (VANET) structure is required

to reduce the number of accidents considerably. Pre-danger information must be communicated in

real-time, in order to implement preventive measures to avoid accidents. VANETs are specially

designed in order to communicate information about hazards. The protocol

DSRC/WAVE/IEEE802.11p is proposed for VANETs, but it may not be enough to cope with

increasing network traffic, especially emergency messages. Hence the use of cognitive Radio (CR)

technology has been introduced. A lot of methods for coordination and channel allocation in the

context of VANETs are being introduced. As such, the need of a framework to reliably compare

the relative performances of different channel sensing, allocation and coordination schemes which

take into account the movement of vehicles is felt. Different techniques like Independent Spectrum

sensing and various forms of Cooperative techniques have been proposed in the near past.

In this dissertation, a VANET structure has been proposed for highways and urban environments.

In the Highway model, separate Road Side Units have been provided for the traffic on each side. In

this way group formation for localized traffic will be easy on highways. Simulation results show

that by using the proposed model, average throughput and end-to-end delay have improved

considerably, while packet loss has also been reduced.

We also propose an efficient spectrum sensing mechanism for sensing and sharing the CR

spectrum by mobile vehicles, which combines best of stand-alone sensing and cooperative sensing

techniques. The proposed mechanism not only improves the probability of correct detection, but

also almost eliminates the probability of misdetection. Then we have introduced a framework that

can be used to define and compare such schemes in a variety of scenarios. Simulation results

clearly show the robustness of our technique by almost eliminating the misdetections and reducing

to a great extent the false alarms. Afterwards we have proposed a sensing technique which prepares

a database for small road segments, time slots for the hours of the day and different frequencies of

the spectrum based on the sensing of vehicles throughout the day. Based on this database, the

future utilization of the spectrum is proposed. Simulations and results clearly indicate the success

and usefulness of our proposed technique.

In the end we have proposed a model based on fuzzy logic for the allocation of different types of

TV channels having different ON/OFF timings in different hours of the day taking time, vehicle

speed, message priority and CR channel sensing results as input. The simulations performed show

the utilization of every type of channel in speed versus time and message priority versus time.

These results can be utilized well for the allocation patterns of CR channels.

Copyright by

SHAHID HUSSAIN ABBASSI

2016

All rights reserved. No part of the material protected by this copyright notice may be reproduced or

utilized in any form or by any means, electronic or mechanical, including photocopying, recording

or by any information storage and retrieval system, without the permission from the author.

DEDICATED TO

My parents, wife and children

CERTIFICATE OF APPROVAL FROM SUPERVISOR

It is certified that the research work contained in this Ph.D. dissertation has been carried out

under my supervision in the Department of Electrical Engineering, Air University, Islamabad. It

is based on original work carried out by the student individually and has not been submitted for

any other degree anywhere else. Moreover, all the other requirements mentioned in the road map

of PhD have been completed. The thesis has also undergone plagiarism test using Turnitin. Its

similarity index is _________________.

Signature: _____________________

Supervisor:

Prof. Dr. Ijaz Mansoor Qureshi

Department of Electrical Engineering

Air University,

Islamabad.

i

LIST OF PUBLICATIONS

1. S. H. Abbassi, I. M. Qureshi, H. Abbasi and B. R. Alyaie, “History Based Spectrum

Sensing in CR-VANETs,” EURASIP Journal on Wireless Communication and

Networking 2015, 2015:163.(ISI indexed, impact factor 0.72)

2. S. H. Abbassi, I. M. Qureshi, B. R. Alyaei, H. Abbasi and K. Sultan, “An Efficient

Spectrum Sensing Mechanism for CR-VANETs,” Journal of Basic and Applied Scientific

Research, vol. 12, no. 3, pp. 365-378, 2013. (ISI indexed)

3. S. H. Abbassi, I. M. Qureshi and H. Abbasi, “Performance of Uni-directional Road Side

Units in Vehicular Adhoc Networks,” in World Congress on Computer Applications and

Information Systems, Hammamet, 2014.

4. S. H. Abbassi, “Algorithm for Topology Discovery in Resilient Packet Ring

IEEE802.17” in SETIT 2009, Hammamet, Tunisia. 22 to 26 March 2009.

5. S. H. Abbassi, “Topology Discovery in Resilient Packet Ring Technology” in INC 2004,

Plymouth, UK. June 2004.

6. S. H. Abbassi, Faheem Ahmed, Mohammad Saeed, “Induction of Survivability into

Rational Unified Process” in SERP 2003, Las Vegas, USA. June 2003.

List of Submitted Paper

1. H. Abbasi, S. H. Abbassi and I. M. Qureshi, “A framework for the simulation of CR-

VANET channel sensing, coordination and allocation,” Journal of Ad Hoc & Sensor

Wireless Networks, Submitted, 2014.(ISI indexed, impact factor 0.478)

The material presented in this dissertation is based on the published papers 1 to 3 and the

submitted paper No. 1.

ii

ACKNOWLEDGMENTS

Thanks to almighty Allah whose blessings have encouraged and provided me strength to conduct

this research and to complete this dissertation. There have been moments when I felt it

impossible to complete my research but almighty Allah has always shown me the way how to do

it.

I am extremely thankful to my supervisor Dr. Ijaz Mansoor Qureshi whose continuous guidance,

and support made it possible to complete this dissertation. His pushing attitude and

encouragement was the key factor throughout my course and research work.

I am highly thankful to Dr. Fida Mohammad Khan whose fatherly attitude provided me help and

moral support. I am also thankful to Dr. Syed Ahmed Pasha whose guidance has helped me to

correct my mistakes.

I am grateful to Mr. Bahman R. Alyaie who’s moral as well as support in problem formulation

and solutions has enabled me to complete the work. I have to give strong credit to my son Mr.

Hameer Abbasi whose help in simulating the problem was a key factor throughout my work. I

am extremely thankful to my wife Ms. Paras Abbasi and my son Janib Abbasi whose handling of

household affairs throughout my research work has enabled me to complete the dissertation.

February 17, 2016

Shahid Hussain Abbassi

iii

TABLE OF CONTENTS

LIST OF PUBLICATIONS ................................................................................................................ i

ACKNOWLEDGMENTS .................................................................................................................. ii

LIST OF TABLES ............................................................................................................................ vii

LIST OF FIGURES ........................................................................................................................ viii

LIST OF ABBREVIATIONS ........................................................................................................... xi

LIST OF SYMBOLS ........................................................................................................................ xv

Chapter 1 INTRODUCTION ......................................................................................... 1

1.1 VANETs ........................................................................................................................... 1

1.2 Cognitive Radio and VANETs ......................................................................................... 2

1.3 contributions of thesis ...................................................................................................... 4

1.4 Organization of the Thesis ............................................................................................... 5

Chapter 2 BACKGROUND ............................................................................................ 6

2.1 MANETs .......................................................................................................................... 6

2.2 VANET ............................................................................................................................ 8

2.3 Security Issues in VANETS ............................................................................................. 9

2.3.1 Communication Patterns ......................................................................................... 10

2.3.2 Overview of Attacks in VANET............................................................................. 10

2.4 IEEE 802.11p / WAVE / DSRC..................................................................................... 12

2.5 Issues in VANETs .......................................................................................................... 15

2.5.1 Vehicle Density ....................................................................................................... 15

2.5.2 High Mobility.......................................................................................................... 16

iv

2.5.3 Intermittent Connectivity ........................................................................................ 16

2.5.4 Definition of Services ............................................................................................. 16

2.5.5 Identification of Service Recipients ........................................................................ 16

2.5.6 Incremental Deployment of VANET ...................................................................... 17

2.5.7 Open Approach to VANET Architecture ............................................................... 17

2.5.8 Unreliable Components Generate Unreliable Data................................................. 17

2.5.9 Privacy .................................................................................................................... 17

2.5.10 Authentication ......................................................................................................... 17

2.5.11 Non Repudiation ..................................................................................................... 18

2.5.12 Reliability, Integrity and Scalability ....................................................................... 18

2.5.13 Real Time Guarantees ............................................................................................. 18

2.6 CR-VANETS ................................................................................................................. 19

2.6.1 Standalone CR Sensing ........................................................................................... 20

2.6.2 Centralized and Cooperative CR Sensing ............................................................... 21

2.6.3 Detection Techniques and Fading Models .............................................................. 24

2.6.4 Spectrum Management and QoS Support ............................................................... 25

2.6.5 Distance Segmentation............................................................................................ 26

2.7 Fuzzy Logic .................................................................................................................... 26

Chapter 3 STRUCTURE OF VANETs ....................................................................... 29

3.1 Message Categorization ................................................................................................. 29

3.1.1 Emergency Message ............................................................................................... 29

3.1.2 Safety Message ....................................................................................................... 30

3.1.3 GPS Message .......................................................................................................... 30

3.1.4 Probe Message ........................................................................................................ 30

v

3.1.5 Traveller Information .............................................................................................. 30

3.1.6 Location Based Service........................................................................................... 30

3.1.7 Informative message ............................................................................................... 30

3.1.8 E-mails .................................................................................................................... 31

3.2 Proposed VANET Structure ........................................................................................... 31

3.2.1 Simulation Results .................................................................................................. 33

3.3 Proposed Highway Structure .......................................................................................... 35


3.4 Summary ........................................................................................................................ 38

Chapter 4 COGNITIVE RADIO AND VANETs ...................................................... 39

4.1 Proposed Spectrum Sensing Framework........................................................................ 39

4.1.1 Network Model ....................................................................................................... 39

4.1.2 Spectrum Sensing Model ........................................................................................ 40

4.1.3 Vehicle Mobility Model .......................................................................................... 41

4.1.4 Spectrum Sensing and Coordination Framework ................................................... 42

4.1.5 Coordinators Selection and Sensing Algorithms .................................................... 47

4.1.6 Simulation and Results ........................................................................................... 52

4.2 Modified Spectrum Sensing and Allocation Model ....................................................... 57

4.2.1 Models Used for the Simulation ............................................................................. 63


4.3 Summary ........................................................................................................................ 71

Chapter 5 HISTORY BASED SPECTRUM SENSING AND ALLOCATION .... 73

5.1 History Updating ............................................................................................................ 74

5.2 History Preservation ....................................................................................................... 76

vi

5.3 History Utilization .......................................................................................................... 77

5.4 Simulation Results.......................................................................................................... 78

5.5 Summary ........................................................................................................................ 89

Chapter 6 CR CHANNEL ALLOCATION USING FUZZY LOGIC IN VANETs

.............................................................................................................................................................. 90

6.1 System Model ................................................................................................................. 90

6.1.1 Inputs and Membership Functions .......................................................................... 91

6.1.2 Outputs .................................................................................................................... 93

6.1.3 Fuzzy IF-THEN Rules ............................................................................................ 96

6.2 Simulation and Results ................................................................................................... 97

6.2.1 Utility Speed Vs Time ............................................................................................ 98

6.2.2 Utility Message Priority Vs Time ......................................................................... 101

6.3 Summary ...................................................................................................................... 104

Chapter 7 CONCLUSION AND FUTURE PROSPECTS ..................................... 105

7.1 Conclusion .................................................................................................................... 105

7.2 Future Prospects ........................................................................................................... 107

Appendix A ...................................................................................................................................... 108

NS2 .................................................................................................................................. 108

C# ..................................................................................................................................... 109

o Visual Studio ................................................................................................................ 109

o C# History and features ................................................................................................ 110

o .NET Framework .......................................................................................................... 111

Bibliography…………………………………………………………………………………...111

vii

LIST OF TABLES

Table 3.1: Comparison 20 to 50 nodes, RSU to DRSU 35

Table 3.2: Comparison 50 to 100 nodes, RSU to DRSU 38

Table 5.1: reply to 76

viii

LIST OF FIGURES

Figure 1-1: An example of VANET ............................................................................................... 2

Figure 2-1: Simple Mobile Ad-hoc Network .................................................................................. 7

Figure 2-2 Modes of Operation in MANETs .................................................................................. 7

Figure 2-3: IEEE 802.11p/WAVE Protocol Stack ....................................................................... 13

Figure 2-4: DSRC 75 MHz Spectrum ........................................................................................... 14

Figure 2-5: IEEE 802.11p Access Layer ...................................................................................... 14

Figure 2-6: Typical Structure of CR_VANET.............................................................................. 19

Figure 3-1: Overall Proposed Structure of VANETs .................................................................... 32

Figure 3-2: Cluster Formation for Bidirectional Traffic ............................................................... 33

Figure 3-3: Cluster Formation for Unidirectional Traffic............................................................. 33

Figure 3-4: Throughput (CASE-1, 50 Mobile nodes). .................................................................. 34

Figure 3-5: Throughput (CASE-2, 50 Mobile nodes). .................................................................. 35

Figure 3-6: Proposed Highways Structure for VANETs .............................................................. 36

Figure 3-7: Throughput (RSUs catering bi-directional traffic)..................................................... 37

Figure 3-8: Throughput (DRSUs catering unidirectional traffic). ................................................ 37

Figure 4-1: Coordinators Selection ............................................................................................... 50

Figure 4-2: Front Coordinator Selection ....................................................................................... 51

Figure 4-3: Back Coordinator Selection ....................................................................................... 51

Figure 4-4: Spectrum Sensing ....................................................................................................... 52

Figure 4-5: Probability of Correct Detection versus No. of Channels .......................................... 54

Figure 4-6: Probability of Correct Detection versus No. of Vehicles........................................... 54

Figure 4-7: Probability of Correct Detection versus Vehicle Velocity......................................... 55

Figure 4-8: Probability of Misdetection versus No. of Channels ................................................. 56

Figure 4-9: Probability of Misdetection versus No. of Vehicles .................................................. 56

Figure 4-10: Probability of Misdetection versus Vehicle Velocity .............................................. 57

Figure 4-11: The Overall Algorithm ............................................................................................. 58

Figure 4-12: Allocation, coordination and sensing ....................................................................... 59

ix

Figure 4-13: How to perform PU state changes. ........................................................................... 60

Figure 4-14: Sensing and occupation mechanism. ....................................................................... 62

Figure 4-15: Allocations Rate vs Vehicles ................................................................................... 66

Figure 4-16: False Alarms vs Vehicles ......................................................................................... 67

Figure 4-17: Misdetections vs Vehicles ........................................................................................ 67

Figure 4-18: Allocations Rate vs Channels .................................................................................. 68

Figure 4-19: False Alarms vs Channels ........................................................................................ 68

Figure 4-20: Misdetections vs Channels ....................................................................................... 69

Figure 4-21: Allocations Rate vs Speed........................................................................................ 69

Figure 4-22: False Alarms vs Speed ............................................................................................. 70

Figure 4-23: Misdetections vs Speed ............................................................................................ 70

Figure 5-1: Segment distribution in a cluster ................................................................................ 74

Figure 5-2: Clusters in a RSU based Network .............................................................................. 75

Figure 5-3: Contents of Tk time slot ............................................................................................. 76

Figure 5-4: Database maintained by the RSUs ............................................................................. 76

Figure 5-5: Data Collection and Sensing ...................................................................................... 79

Figure 5-6: Data Calculation ......................................................................................................... 80

Figure 5-7: History Records Updating .......................................................................................... 81

Figure 5-8: Allocation Rate Vs No. of Cars ................................................................................. 83

Figure 5-9: Allocation Rate Vs No. of Channels .......................................................................... 84

Figure 5-10: Allocation Rate Vs Speed ........................................................................................ 84

Figure 5-11: False Alarm Rate Vs No. of Cars ............................................................................. 85

Figure 5-12: False Alarm Rate Vs No. of Channels ..................................................................... 85

Figure 5-13: False Alarm Rate Vs Speed ..................................................................................... 86

Figure 5-14: Rejection Rate Vs No. of Cars ................................................................................. 86

Figure 5-15: Rejection Rate Vs No. of Channels ......................................................................... 87

Figure 5-16: Rejection Rate Vs Speed .......................................................................................... 87

Figure 5-17: Forced Leave Ratio Vs No. of Cars ......................................................................... 88

Figure 5-18: Forced Leave Ratio Vs No. of Channels.................................................................. 88

Figure 5-19: Forced Leave Ratio Vs Speed .................................................................................. 89

x

Figure 6-1: Overall Fuzzy Logic System ...................................................................................... 90

Figure 6-2: Membership Function of Speed ................................................................................. 91

Figure 6-3: Membership Function of Message Priority ................................................................ 92

Figure 6-4: Membership Function of Time .................................................................................. 92

Figure 6-5: Availability of Channel-1........................................................................................... 94





Figure 6-10: Availability of Channel-6......................................................................................... 95



Figure 6-13: Speed Vs Time for Channel 1 .................................................................................. 98





Figure 6-18: Speed Vs Time for Channel 6 ................................................................................ 100



Figure 6-21: Message Priority Vs Time for Channel 1............................................................... 101



Figure 6-24:Message Priority Vs Time for Channel 4................................................................ 102





xi

LIST OF ABBREVIATIONS

MANETS Mobile Ad-hoc Networks

AP Access Point

BSS Basic Service Set

VANETS Vehicular Ad-hoc Networks

NHTSA National Highways Traffic Safety Administration

RSU Road Side Unit

V-I Vehicle to Infrastructure

V-V Vehicle to Vehicle

NOW Network on Wheels Group

PATH Partners for Advanced Transit and Highways

DSR Dynamic Source Routing

AODV Ad-hoc on Demand Distance Vector

DSDV Destination Sequenced Distance Vector

DoS Denial of Service

VIN Vehicle Identification Number

WAVE Wireless Access in Vehicular Environment

DSRC Dedicated Short Range Communication

EU European Union

HTTP Hyper Text Transfer Protocol

TCP Transfer Control Protocol

UDP User Datagram Protocol

IPv6 Internet Protocol version 6

xii

LLC Logical Link Layer

FCC Federal Communication Commission

CCH Control Channel

HALL High Availability Low Latency

SCH Service Channel

QoS Quality of Support

EDCA Enhanced Distributed Coordinated Access

MAC Media Access Control

CSMA/CA Carrier Sense Multiple Access with Collision Avoidance

DCF Distributed Coordination Access

FHSS Frequency Hopping Spread Spectrum

DSSS Direct Sequence Spread Spectrum

IR Infra-Red

OFDM Orthogonal Frequency Division Multiplexing

HR High Rate

CCK Complementary Code Keying

PBCC Packet Binary Convolutional Code

DIFS Distributed Inter Frame Spacing

NAV Network Allocation Vector

CW Contention Window

AC Access Category

AIFS Arbitrary Inter Frame Space

CCI Control Channel Interval

CR Cognitive Radio

xiii

GSM Global System for Mobile Communication

CDMA Code Division Multiple Access

LTE Long Term Evolution

LAPUs Local Acquisition and Processing Units

ISM Industrial, Scientific and Medical

PU Primary User

CBS Cognitive Base Station

MAP Maximum a Posteriori

RAT Radio Access Technologies

AF Amplify and Forward

SU Secondary User

VDSA Vehicular Dynamic Spectrum Access

CCC Common Control Channel

Cog-V2V Cognitive Vehicle to Vehicle

FL Fuzzy Logic

PSO Particle Swarm Optimization

FUZZBR Fuzzy Logic based multi-hop Broadcast Protocol

NS2 Network Simulator version 2

C# See Sharp

MATLAB Matrix Laboratory

NAM Network AniMator

VINT Virtual InterNetwork Testbed

DARPA Defense Advanced Research Projects Agency

OTcl Object Tool Command Language

xiv

IDE Integrated Development Environment

GUI Graphic User Interface

CLI Command Language Infrastructure

CLR Common Language Runtime

LINPACK LINear Equations Software Package

EISPACK Eigensystem Package Subroutine Computing Facility

LCA Legal Certification Authority

RCAs Regional Certification Authorities

SPs Service Providers

VCA Verification and Controlling Authority

DRSUs Directional Road Side Units

RCUs Road Central Units

GPS Global Positioning System

VHF Very High Frequency

UHF Ultra High Frequency

CATV Community Access Television

SNR Signal to Noise Ratio

ACK Acknowledgement

NACK Non- Acknowledgement

MIMO Multi Input Multi Output

xv

LIST OF SYMBOLS

{ }

Center Frequency of individual channel

CH HALL Channel of DSRC Spectrum

Pilot frequency of primary user networks ( indicates the number of

channel and * indicate the technology used, v for VHF, u for UHF, and t

for CATV, s for secondary user)

H1 Presence of Primary user signal

H0 Absence of Primary user signal

Ci CR Channel number

( ) Detected signal

Re{.} Real part of the complex waveform

( ) Equivalent low pass representation of the detected primary or secondary

user signal.

( ) Additive white Gaussian noise with zero mean

Energy of the detected signal

Maximum time period

( ) Probability density function

Gamma function

I Modified Bessel function

Instantaneous signal to noise ratio (SNR)

Degree of freedom

( ) Safe speed

xvi

( ) S peed of leading vehicle at time t

( ) Gap of leading vehicle at time t

Average speed

( ) Deceleration function

Driver’s reaction time (usually 1 second)

Maximum allowable speed of the vehicle

Acceleration

( ) Desired speed

Human error factor between 0 and 1

( ) Final speed at time t

Main coordinator

Forward edge coordinator

Backward edge coordinator

Request

Request categories

Cluster member vehicle

{ ( )} Likelihood of channel availability

Probability of detection

Probability of detection given

Instantaneous SNR of different (VHF, UHF, CATV) signals on CR

spectrum

Average SNR

Result Array. * represents M for Main, F for Forward and B for Backward

Coordinators

Median path loss in urban environment

xvii

Base station antenna height in meters

Mobile station antenna height

Carrier frequency in MHz

Antenna height correction factor

Distance in kilometers

Predicted median path loss for suburban areas in decibels

Rayleigh random variable

Mode of the distribution

Distance segment

Third last segment of backward cluster

Front coordinator of backward cluster

Third segment of front cluster

Backward coordinator of front cluster

M No. of channels

Sensing interval

Vehicle identity,

X and Y Coordinates received from GPS to locate the segment

CH Channel sensed from 1 to M

RES Sensing result (1 or 0)

Segment number

K No. of time slots

Time slot number

S No. of time intervals in one

Probability of availability

xviii

Ratio of historic data.

1

Chapter 1

INTRODUCTION

1.1 VANETS

Approximately 1.24 million people die every year and nearly 20 to 50 million are affected by

fatal injuries due to road accidents around the globe according to a report published by the World

Health Organization [1]. The report further states that road accidents are the eighth leading cause

of fatal injuries and may become the fifth leading cause if proper measures are not taken to

reduce road accidents. Road traffic injuries are estimated to cost over US$ 100 billion to low and

middle income countries, which is an estimated 1-2% of their gross annual product [2]. Report

by National Highway Traffic Safety Administration (NHTSA) also finds out the fatalities in road

accidents from 1994 to 2012 [3].

Vehicular Ad-hoc Networks (VANETs) are a special form of MANETs implemented on road

infrastructure, where the mobile nodes are vehicles. Road side units (RSUs) are installed at a

particular distance along the road and road junctions for vehicle to infrastructure (V-I)

communication and for vehicle to vehicle (V-V) communication for vehicles at larger distance

apart and hence cannot communicate directly. VANETs differ from MANETs due to high speed

of vehicles, high battery power, high density, intermittent connectivity, security issues, limited

vehicles movement, static geometry of road, reliability, integrity, large scale infrastructure, real-

time guarantees.

VANETs are especially designed to decrease and then eventually eliminate the road accidents

while increasing the travel comforts. Road accidents are decreased by timely inter-vehicle

communication to inform the vehicles behind about e.g road broken, road blocked, railway

crossing, bridge broken, junction ahead, security barrier, and accident warning. Travel comfort is

2

enhanced by providing gaming services, Internet, fuel station locations, restaurant locations,

service area locations, and on-line payments [4]. Figure 1-1 provides a view of a normal VANET

[5].

Figure 1-1: An example of VANET

Besides IEEE several consortiums are working on the development of VANETs. These include

Car-to-Car consortium and Network on Wheels group (NOW) in Europe; Berkley PATH in

USA; and Fleetnet Projects in Germany [6] [7].

1.2 COGNITIVE RADIO AND VANETS

Due to increase of vehicular traffic resources provided for the vehicular communication are

becoming insufficient. Especially emergency messages require reliable and fast enough

communication resources in order to avoid accidents. Provision of communication by Cognitive

Radio (CR) technology is beneficial for vehicular communication for which accurate spectrum

sensing and coordination techniques are needed.

3

A lot of wireless technologies are in practice around the globe today, like Wi-Fi, Wi-Max,

Cellular Telephone Technologies (e.g. GSM, CDMA, and LTE), TV bands, and IR remotes. It

can be found by scanning that some of the frequency bands are over utilized, while others are

under-utilized which costs a lot to the authorities. CR is an excellent way of dealing with the

spectrum under-utilization problem. It can be observed by means of scanning that some of the

slots of wireless spectrum are occupied by the primary (Licensed) users, while some slots are

empty. Empty slots can be made available to secondary (un-licensed) users which are denied

channel access due to congestion in their own spectrum. Empty slots are referred to as holes [8]

[9]. There are three transmission modes of CR, which are interweave, spectrum overlay and

spectrum underlay. Comparison of these modes is discussed in [10].

Many signal detection techniques can be used for spectrum sensing in order to improve the

probability of detection. Some techniques are energy detection, matched filter detection, cyclo-

stationary detection, and wavelet detection. Every technique has its own pros and cons [11] [12].

Spectrum sensing is a very important aspect of cognitive radio networks as the important

decision of using a licensed spectrum is to be taken based on the results obtained by spectrum

sensing. Hence spectrum sensing must be carried out accurately and in a timely manner so that

reuse of the licensed spectrum by the secondary users can be achieved with minimum or even

negligible interference to the licensed primary users [13] [14].

In CR, unlicensed (Secondary) users are allowed to borrow unutilized bandwidth from licensed

(Primary) users. Sharing of network resources continues in this fashion until primary users need

more of the spectrum, in which case secondary users have to vacate the spectrum [15] [16].

Cognitive Radios can be built so they are smart enough to use parameters (such as carrier

frequency, bandwidth, and transmission power) of that particular spectrum band and particular

wireless technology [17].

It can be observed that some spectrum bands like TV channels have spatio-temporal patterns.

During some hours of the day and in some areas some channels are on while others are off.

These patterns can be exploited to improve the efficiency of spectrum sensing by making it more

directed [18].

4

1.3 CONTRIBUTIONS OF THESIS

This dissertation highlights several issues related to the VANETs and use of cognitive radio

technology in VANETs. Main objective of this work is to eliminate the mis-detections and

reduce the false alarms in CR-VANETs. Main contributions of this thesis are summarized as

follows:

Initially a structure of VANETs is proposed for highways and urban environment and

then emphasises is given to highway portion of the structure and more simulations are

conducted to verify its worth. It is proposed to use Directional Road Side Units (DRSUs)

instead of normal Road Side Units (RSUs).

Cognitive radio concept is used in VANETs. We have proposed multi-spectrum sensing

algorithm based on three coordinators to sense and allocate un-occupied channels to

secondary users (vehicles). Results were compared with other popular approaches.

The proposed algorithm was modified and in order to create realistic environment, Hata

model [19] for large scale fading and Rayleigh Fading model [20] for small scale fading

were used. Also modification in Gibbs Mobility model has been proposed for the

movement of vehicular traffic [21].

This algorithm was further enhanced to build and periodically update the history of

cognitive sensing results during the day time in order to utilize this history to find the

most probable holes for use by the vehicles. The database is stored at the DRSU and

given to the requesting vehicles through the coordinators. Hence bulk of sensing is

decreased by utilization of the historic sensing database built on the basis of spatio-

temporal-frequency [22].

Fuzzy Logic techniques are used for the allocation of CR channels for the vehicular

network. On/Off timings of different types of TV channels, speed of the vehicles,

message priorities and time dividing twenty four hours of the day are considered as inputs

to fuzzy engine to find out the allocation of TV channels to vehicles.

5

1.4 ORGANIZATION OF THE THESIS

This thesis has been organized as follows. Chapter 2 provide the background over which our

work is based. Chapter 3 includes our proposed ‘Structure of VANETs’. Chapter 4 provide the

details of our proposed techniques for the use of cognitive radios in VANETs. Chapter 5 includes

our work involving use of historic sensing data for allocating CR channels to vehicular nodes.

Chapter 6 includes our work to use fuzzy logic techniques for the allocation of cognitive radio

channels to the vehicular nodes. Chapter 6 includes the conclusion and future work. Appendix A

provide the details of tools used for simulations for our entire work.

6

Chapter 2

BACKGROUND

2.1 MANETS

Mobile Ad-hoc Network (MANET) is a wireless temporary network which may change

continuously with the entrance and exit of mobile nodes. In MANETs mobile nodes form an

autonomous transitory association by communicating among themselves using a wireless

medium. MANETs do not rely on the fixed infrastructure for communication instead information

is shared through wireless nodes which may be single or multi-hop. Nodes which lie in each

other’s communication range dynamically discover each other and communicate on single hop

basis and the nodes which are not in the communication range of each other rely on the

intermediate nodes to act as routers to relay the packets to the destination node and communicate

on multi-hop basis [23] [24] [25]. As these nodes are battery powered, so these are often energy

constrained. These nodes randomly move around resulting in rapid and un-predictable

topological changes. In such an un-predictable multi-hop environment nodes have to act quickly

to form a network and relay packets in the absence of access points (fixed infrastructure). Figure

2-1 shows a simple example of MANET [26].

7

Figure 2-1: Simple Mobile Ad-hoc Network

Considering IEEE 802.11 as MAC (Medium Access Control) layer MANETs have two modes of

operation. One is ‘Infrastructure mode’ and another ‘Ad-hoc mode’. In Infrastructure mode all

nodes communicate with each other via an ‘Access Point (AP)’ which is connected to the

internet through wired media. AP behaves as a base station for all nodes. Each AP has a

configurable communication range termed as ‘Basic Service Set (BSS)’. In order to extend the

range more APs can be installed. There need to be an overlap among the cells of APs to avoid

breaks in the communication. In Ad-hoc mode all the users within certain area form a personal

network among each other. Every intermediate node behaves as a router in case distance between

transmitter and receiver is greater than their communication range. Figure 2-2 gives a pictorial

view of both modes [26].

Figure 2-2 Modes of Operation in MANETs

8

Major characteristics of MANETs include; autonomous, infrastructure-less and less expensive. It

is scalable and more flexible. It involves multi-hop routing, dynamic network topology and

device heterogeneity. MANETs are self-created, self-organizing and self-administrable

networks. [27] [28]:

Some complexities and challenges of MANETs include; Lower reliability compared to wired

medium, dynamic topology, routing overhead, limited physical security, time varying channels,

hidden terminal problem, packet losses due to transmission errors, mobility induced route

changes, battery constraints, security threats, interference, energy constrained operation,

bandwidth constrained variable capacity multi-hop links [27] [28].

Some of the MANET applications include; military battlefield, collaborative work, local level,

personal area networks, commercial sectors, emergency services, education, entertainment,

coverage extension [27] [28].

2.2 VANET

VANETs are being designed specially to avoid loss of lives due to road accidents of millions of

people around the world. VANETs are also intended to provide toll services, location based

services, and infotainment. Main applications of VANETs are collision avoidance, cooperative

driving, traffic optimization, payment services, location based services, entertainment

applications [29].

Unlike MANET, most VANET nodes (Vehicles) are equipped with large batteries and due to

availability of a charging mechanism within the vehicle; the power problem is almost negligible

[30] [31] [32]. VANETs are specially designed for moving vehicles that include many

applications classified as public safety, traffic management, freight/cargo transport, transit,

traveler information/support etc. The primary goal of the public safety application is to reduce or

even eliminate accidents which result in fewer injuries and fatalities, lowering the direct or

indirect financial costs, and reducing traffic congestion. Examples are forward obstacle detection

and avoidance, lane departure warning, turn accident warning, intersection accident warning, low

9

bridge warning, roll over warning, work zone warning, stopped vehicle warning and railway

crossing warning. The goal of traffic management is to improve the flow of traffic in order to

facilitate the passengers and drivers and reduce travel time. Application examples are smart

traffic signals, variable message signs, rapid response to incidents, enhanced public transit,

emergency vehicle warning, central traffic management and electronic toll collection [33].

Enhanced transit systems include traffic signal priority, bus only lane enforcement, bus turn light

priority, automated fare collection and reporting, automated passenger counting, route

optimization and schedule tracking, rider information, on demand transit services, security

systems, fleet operations and maintenance, parking, and many other on-board systems. Freight

and cargo systems include vehicle registration/ inspection and credentials, route guidance and

tracking, vehicle monitoring and maintenance systems, cargo monitoring and tracking, and fleet

operations. Traveler information and support includes pre-trip planning, route and fare

information, access to news, weather reports and internet, navigation aids, traffic information on

routes, access to personal information during the trip, restaurant and fuelling station information,

and vehicle repair center information. Other entertainment services include audio, video, and

email services [34] [35].

Due to a different nature of VANETs as compared to ordinary MANETs, different routing

protocols are being considered. These include Reactive protocols like Dynamic Source Routing

(DSR), Ad-hoc on Demand Distance Vector (AODV) and Proactive protocols like Destination

Sequenced Distance Vector (DSDV), Wireless Routing Temporally- Ordered Routing

Algorithm, and Lightweight Mobile Routing protocols [36] [32].

2.3 SECURITY ISSUES IN VANETS

Unlike other types of MANETs, loss of human lives can be caused if any incorrect warning or

message is communicated in VANETs. Hence special security measures are needed. Basically

there are two different environments in VANETs: Infrastructure environment and Ad-hoc

environment. Legal authorities, manufacturers, trusted third parties and service providers form

the Infrastructure environment which provides the services to vehicles through RSUs. Vehicle to

10

vehicle communication through communication units (On Board Units) installed in vehicles,

form the Ad-hoc environment [29].

2.3.1 COMMUNICATION PATTERNS

Different communication patterns are found in VANETs, which are detailed below [29].

a. V2V Warning Propagation

Messages like road closure or accident warning for the vehicles behind and the messages like

'ambulance or other emergency vehicle coming' in order to get the lane cleared by the preceding

vehicles.

b. V2V Group Communications

Messages for vehicles belonging to a particular area or the vehicles of particular sort like

vehicles manufactured by single manufacturer, or all public transport vehicles or all load

carrying vehicles.

c. V2V Beaconing

Messages like current speed of vehicle, direction of vehicle, use of brakes. These messages are

mostly one hop messages delivered to one vehicle behind.

d. I2V/V2I Warning

These include messages like intersection ahead, railway crossing ahead and service area ahead,

fall in the category of Infrastructure to Vehicle (I-V) whereas the messages like road closure fall

in the category of Vehicle to Infrastructure (V-I).

2.3.2 OVERVIEW OF ATTACKS IN VANET

Different sort of attacks are found in VANETs. These are classified as under [29] :

11

a. Attacks on identification and authentication

Two types of this sort of attack are Impersonation and Sybil. In Impersonation, the attacker

pretends to be another entity by stealing the attributes of any other entity (vehicle). In Sybil one

entity poses the credentials of many other entities and behaves as different entities at different

times.

b. Attacks on Privacy

Sometimes, attackers try to obtain the identity of the driver (usually owner) and sometimes they

try to find the location of the vehicle to harm the passengers.

c. Attacks like non-repudiation

Some people try to deny the fact that they have sent any message like wrong road condition

information (road closure, speed, crossing ahead etc.). Also sometimes the people deny the

receipt of any warning message and may cause accidents.

d. Attacks on confidentiality

Sometimes attackers try to listen into others' communications, to gain access to confidential

information (eavesdropping). This type of attack is quite serious and may cause huge losses to

any person or enterprise.

e. Attacks on availability

Availability of any service or hardware like RSU is very important for services to continue.

Some attackers try to hide some services and may cause the overloading of communication

channels (Denial of Service attack or DoS attack). Some attackers overload the computation

capabilities of any vehicle to cause loss to the vehicle or passengers.

f. Attacks on data trust

Some attackers create inaccurate data to harm others. This may reduce the reliability of the

whole system.

For security purposes different mechanisms have been proposed and provided. Manufacturers

issue each vehicle a vehicle identification number (VIN), whereas legal authorities issue License

12

Plate number. VIN is assigned to uniquely identify manufactured vehicle, while License Plate

number uniquely identifies the vehicle in an administrative domain.

To deal with the issue of authentication 'Electronic License Plate' has been proposed which will

not only identify the vehicle but also authenticate it. For the purpose of privacy Public Key

Certificates have been proposed. Two techniques being used for this purpose are Identity based

Cryptography and Pseudonymous Short-lived Public Key Certificates. [37]

For the location of any vehicle by legal authorities, a Location Cloaking Technique [38] has been

provided. Aggregation technique [39] has also been used for this purpose.

2.4 IEEE 802.11P / WAVE / DSRC

A special frequency band of 5.850-5.925 GHz has been allocated for the purpose of vehicular

communication in USA [40]. Similar bands have also been allocated in Japan and Europe. Due

to high mobility normal IEEE-802.11 is not suitable for VANET applications so IEEE has

developed a special IEEE-802.11p standard for VANETs [41] [7] [42]. IEEE 802.11p/WAVE is

not a single standard, but it is a group of standards which comprises P1609.1 (Resource

Manager), P1609.2 (Security Services), P1609.3 (Networking Services), 1609.4 (Multi-channel

Operations) as shown in Figure 2-3, and 802.11p (MAC Layer) [36] [43] [44]. The European

Union (EU) is also getting close to allocating 30 MHz in 5 GHz band, especially for the

vehicular communication [45].

13

Figure 2-3: IEEE 802.11p/WAVE Protocol Stack

The IEEE has standardized 802.11p/WAVE (Wireless Access in Vehicular Environment) for

vehicle to vehicle communication(V-V) and vehicle to infrastructure communication (V-I). This

process has been initiated with DSRC (Dedicated Short Range Communication) spectrum

allocation. The USA Federal Communication Commission (FCC), in the year 1999, allocated 75

MHz, 5.9 GHz band DSRC especially for the V-V and V-I communication as shown in figure 2-

4. As shown in the figure, DSRC is divided into seven 10 MHz channels. Channel 178 is control

channel (CCH) and is especially reserved for public safety messages. Channel 172 is High

Availability Low Latency (HALL) channel and is reserved for critical safety V-V applications.

Channel 184 is reserved for High power public safety applications. The other four channels are

service channels (SCH) used either for safety or non-safety applications [13]. Channels 174 and

176 or 180 and 182 can be combined to get two 20 MHz channels. The range of the DSRC

standard is around 300 meters to a maximum of 1000 meters. Its data rate is from 6 to 27 Mbps.

It is half-duplex communication standard.

14

Figure 2-4: DSRC 75 MHz Spectrum

IEEE 802.11p uses physical layer of IEEE 802.11 and utilizes the advanced QoS support

provided by IEEE 802.11e as shown in Figure 2-5 [46]. It utilizes the Enhanced Distributed

Coordinated Access (EDCA) feature of 802.11e QoS in the MAC layer, which is based on

Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) also known as

Distribution Coordination Function (DCF) in IEEE 802.11.

Figure 2-5: IEEE 802.11p Access Layer

In DCF, the station will only transmit a frame if it senses the medium free for a certain time

period known as Distributed Inter Frame Spacing (DIFS). Another time period known as

Network Allocation Vector (NAV) is also defined in IEEE 802.11 which specifies that for how

long the medium is busy. Stations must wait for NAV to complete before transmitting a frame.

Another timer is Contention Window (CW). If there is a collision; station start a back-off timer

15

between CWmin to CWmax and after the timer stops; stations start sensing the medium and

sending the frames.

In EDCA MAC some improved features have been introduced which include four priority

queues for Background, Best Effort, Video, and Voice traffic, which are called Access

Categories (AC). DIFS interval in EDCA is called Arbitrary Inter Frame Space (AIFS). Each AC

has different CWmin and CWmax values based on priorities and AIFS is calculated based on

different CW value. Voice traffic has highest priority and AIFS calculated for this AC is around

34 µs as compared to 41 µs DIFS in IEEE 802.11a. In case of collision higher AC wins access to

the medium.

The nature of traffic is different on highways as compared to urban environments. Hence a

flexible structure is required to address the issues of both types of road networks. Furthermore

there are several issues that may arise while designing the structure and standards for VANET.

Categorization of messages is also needed to save the bulk of overhead due to security in

message size.

2.5 ISSUES IN VANETS

VANETs face several issues which need to be addressed while designing the standards and a

foolproof structure for VANETs [4] [47]. These issues are described as follows:

2.5.1 VEHICLE DENSITY

In urban areas road network face a high density of vehicles and hence vehicles are bound to move

with slower speed, whereas there is low density of vehicles on highways and motorways allowing

the vehicles to move with higher speed. Therefore communication network design for the urban

areas needs to be treated differently than the highways considering Transmission, Routing,

Quality of Service, Security and Location [48].

16

2.5.2 HIGH MOBILITY

In normal MANETs nodes move with slower speed as compared to the Ad-hoc network formed

by the vehicles. Hence in MANETs Ad-hoc network remain intact for a larger duration compared

to that in vehicular network. Therefore VANETs are treated differently than MANETs especially

while designing the MAC layer [49].

2.5.3 INTERMITTENT CONNECTIVITY

In the vehicular network the probability of connectivity between the vehicles moving in opposite

direction is much smaller compared to the vehicles moving in same direction. Road side

Infrastructure Units (RSUs) can be installed to avoid the intermittent connectivity among the

vehicles [50].

2.5.4 DEFINITION OF SERVICES

Network infrastructure on roads for vehicles involves different type of services when compared to

normal MANETs. These services include toll collection, emergency services like ambulances,

traffic police, location based services including restaurants, fuel stations, workshops. It also needs

the fast dissemination of safety messages which include safe distance between the vehicles,

accident warnings, road closure warnings, and road congestion warnings [51] [52].

2.5.5 IDENTIFICATION OF SERVICE RECIPIENTS

There are some services which involve online payments such as toll payments, online shopping.

In these types of services actual service recipient, who has paid for the service, needs to be

identified accurately [53].

17

2.5.6 INCREMENTAL DEPLOYMENT OF VANET

As the research on VANETs is gaining pace, new ideas are pouring in faster day by day.

Therefore network design shall be more flexible in order to accommodate new changes

facilitating the passengers and drivers [53].

2.5.7 OPEN APPROACH TO VANET ARCHITECTURE

Around the globe in different continents different approaches of VANET have been applied. This

situation is causing non-uniformity in the networks. It is an ultimate requirement to bring the

uniformity to facilitate users travelling around the globe.

2.5.8 UNRELIABLE COMPONENTS GENERATE UNRELIABLE DATA

A large population on the earth can be trusted and we may find more sincere people but there is

some minority which cannot be trusted and hence can generate unreliable data harmful for the

networks. In VANETs this harmful data may cause accidents. Hence the networks shall be

designed keeping in view the rigidity and the security to avoid the harmful activities [54].

2.5.9 PRIVACY

Personal information and identity of any person on the network is needed to be secretly placed so

that no one can know the location of the person and data being used by him except the authority

controlling the network in order to avoid misuse of data and harm to any individual [55].

2.5.10 AUTHENTICATION

Any person using a network service or generating any sort of safety message is strictly needed to

be authenticated as any sort of false safety message cause road congestion or accidents and hence

loss of lives [56].

18

2.5.11 NON REPUDIATION

Cases may occur that some people may deny receipt of any message or sending of any message.

This case may be harmful if someone denies receipt of messages from controlling authorities like

speed control warnings. Hence attention shall be given to design more reliable systems in order to

avoid such occurences [57].

2.5.12 RELIABILITY, INTEGRITY AND SCALABILITY

The communication channels used by the vehicular networks shall be scalable and reliable so that

message delivery can be guaranteed and top priority shall be given to the message integrity [58].

It has been pointed out in [59] that an increase in the control channel interval (CCI) will increase

the reliability of safety as well as non-safety applications in VANETs.

2.5.13 REAL TIME GUARANTEES

Road warnings and safety messages shall be delivered in real-time, to avoid the mishaps and

hazards on the road [60]. Authors in [59] have mentioned that in IEEE 802.11p, any

acknowledgement for broadcast messages is not sent by the vehicles, hence safety related

messages, just like collision warning, may not be received within short time by the following

vehicles which may cause chain accidents. It has been shown in some simulation based papers

that in IEEE 802.11p as the vehicle density increases, the latency of the packets also increases,

which decreases the successful packet reception rate [61] [62]. [63] [64]. Authors in [61] and

[62] have formed an algorithm to control the load for periodical status messages. [63] analyzes

the channel access delay in DSRC and it has been compared with the self-organizing time-

division multiple-access scheme. It has been proved that later scheme performs with better

results. In [64] a framework has been proposed to share the DSRC between safety and non-safety

applications. In [65] a 1-D Markov model has been proposed to calculate the delay and reception

rate in VANETs. In [66] the average delay in DSRC for each access category (AC) was analyzed

without considering back-off delay.

19

2.6 CR-VANETS

Due to the increase in the number of vehicles, it has been observed that DSRC may not be able to

cope with the large number of messages to be communicated in VANETs. Hence it is evident

that vehicles may have to utilize holes in other licensed spectrums, especially for safety critical

messages. It has been observed that VHF and UHF bands are not fully utilized, hence may be

used by un-licensed users. Considering the critical safety applications of VANETs, accuracy in

spectrum sensing is important. A lot of research work is being carried out on use of cognitive

radio in VANETs. Some have proposed stand-alone sensing schemes [67] [68], and some have

proposed cooperative spectrum sensing [11] [12]. Typical architecture of VANETs including

Cognitive Radios is shown in figure 2-6 [69].

Figure 2-6: Typical Structure of CR_VANET

There may be a gap between the methods used to produce results for such techniques, which

may cause discrepancy in the data. For example, one person may use a different fading model to

the next. Another may include small-scale fading while the first may not, which introduces

additional discrepancy in the results. One may use a different mobility model to the next. One

may use a different channel sensing technique to another. All of them are equally valid, but they

lack a common platform for impartial comparison, which takes into account variables such as

movement, fading, noise, interference [70] and other such factors which may cause variability in

20

the simulation results and at the same time allows for varying parameters, in order to perform

simulations in a variety of conditions. The historic approach in the scientific method has always

been that the data being presented should be falsifiable. [71] These differences introduce a

difficulty as to the falsifiability of the data, and even more importantly, the usefulness of such

techniques as comparison is obscured.

As VANET involves many safety related issues, hence the use of cognitive radios (CRs) in

VANETs have forced researchers around the globe to propose the solid and concrete models best

suited specifically to safety of vehicles and passengers. Many different models have been

proposed for the use of CRs in VANETs emphasizing more on the priority information

exchanged among vehicles and between RSUs and vehicles. Since life of the passengers and

pedestrians will be in danger if any delay is caused in safety critical message or the received

message is wrong; hence special attention needs to be given to real time guarantees and

reliability of data considering high mobility and density of vehicles. In IEEE 802.11p/WAVE,

control channel (CCH) is very important and the load of messages on this channel is very high,

so some researchers have proposed in [72] a structure comprising Local Acquisition and

Processing Units (LAPUs), RSUs and vehicles especially for the use of cognitive radios in TV

bands. LAPUs and RSUs are basically the decision centers, hence this architecture is centrally

controlled by these units. Responsibility of the RSUs is to assign the holes to the requesting

vehicles in their cell (Coverage area of RSUs) considering the load on CCH applied by the

member vehicles.

2.6.1 STANDALONE CR SENSING

In [73] researchers have proposed that a specific channel from TV spectrum band shall be

assigned to each vehicle to sense and use. Every vehicle shall sense and use the spectrum

independently and information gathered be exchanged between vehicles for future decisions on

the use of spectrum. In this proposed scheme there is no coordination among the vehicles and a

particular hole may be occupied by more than one vehicle at a time which may damage the data

of all vehicles using that hole. Hence this problem needs special attention. Researchers in [74]

have proposed unlicensed Wi-Fi channels in urban areas and ISM bands (2.4 GHz or 5 GHz) are

21

not fully occupied by the primary users (PUs), hence the vehicles can periodically sensed, and

use the spectrum holes and share the information between each other for future usage. Another

solution is proposed in [75] based on belief propagation. In this proposal every vehicle

broadcasts a message to its neighboring vehicles containing the information about the presence

of primary user signals in its range. All vehicles receiving this message decide on the basis of

belief messages received from all of their neighbors and their own observations; that which PU

channels to use for communication. As this scheme involves messaging among all the

neighboring vehicles, which creates a lot of communication and suffers from slow processing

due to dependence on all of its neighbor data. In [43] it is proposed that each secondary user

senses the PU spectrum and uses if the hole is detected. It also shares the information on DSRC

channels with its neighbors. In [76] authors have proposed a three state model, one describing

the empty slot (hole), other specifying the presence of primary user, and third specifying the

occupancy of channel by any secondary user. This scheme also insists on individual spectrum

sensing by the secondary users with no collaboration. The proposals presented in [16], [17], [18],

[19], and [20] are based on the decentralized structure and may suffer from congestion,

interference and collisions.

2.6.2 CENTRALIZED AND COOPERATIVE CR SENSING

Cooperative spectrum sensing can be effective in high mobility environment [77] A group of

vehicles may combine to sense and form a centralized database to be used for utilization for a

cluster. Authors have shown in [78] that as a result of spatial temporal diversity received signal

strength can be increased by using cooperative spectrum sensing in high mobility environment.

To reduce the sensing overhead they have proposed an optimal number of sensors to use for

cooperation and number of times to sense. In [79] authors insist to enhance the reliability and

reduce the delay in sensing by the use of cooperative spectrum sensing. They explain that by the

use of spatial diversity the problem of sensing limitations of a single vehicle can be tackled very

well. It is also shown that sensing time can be reduced by the use of cooperative sensing

compared to sensing by a single vehicle. Another work in [80] presents a RSU controlled central

architecture. RSUs sense the spectrum of TV bands and prepare a database to be used by

22

requesting CR-enabled vehicles. In the case centralized infrastructure is not present then

clustering technique can be used with one cluster head responsible for sensing the spectrum and

sharing it with member vehicles on demand [81]. In [82] researchers have proposed to form the

clusters of the vehicles which are moving in groups and one vehicle may be assigned the duty of

cluster head. Spectrum sensing of PU spectrum is performed by every cluster member and results

are sent to the cluster head. It is the responsibility of the cluster head to assign the holes to every

cluster member based on the information received from every cluster member. In this proposal

decision is taken on the basis of hole availability information received from all vehicles and a hole

sensed by one vehicle may be assigned to another vehicle which may be in different environment

due to various factors like shadowing and distance from PU tower. This scheme may also suffer

from delay as the holes detected may be occupied till the decision is received from the cluster

head. In [83] it is proposed to divide the wideband spectrum of CR network in small sub-bands

and assign each small narrow band to a group of nodes (vehicles) to scan and use the spectrum.

Secondary users sense the spectrum and send the results to a Cognitive Base Station (CBS) which

detects the transmitted values using MAP (Maximum a Posteriori) detector and then fuse the

results to find the occupancy of the channel. This scheme may save the time of scanning as

several groups are sensing different narrow bands at a time but lacks the correct decision making

policy as no factor is considered in assigning the narrow band to a group of vehicles. A particular

narrow band may suit one group of vehicles but may not suit another group.

In [84] authors have proposed a cooperative framework for spectrum sensing. Each vehicle senses

the PU spectrum individually and shares the information with neighbors. Each sample of

information collected from a user is assigned a weight, and then the weight is adjusted and

normalized. Afterwards, a decision is made for the availability of a hole and the decision is

forwarded to all neighbors. This paper does not propose the method of utilizing of hole after it is

found. In [85] energy detection techniques have been discussed. This paper does not mention any

cooperative mechanism that vehicles will sense the spectrum and decide about the availability of a

hole and allocation to any secondary user. In [86] multi radio technologies (Multi-RAT) have

been discussed and it has been proposed that every vehicle will sense or use any technology

according to the required class of service. For example it proposes the usage of Wi-MAX for

23

video conferencing and to use Wi-Fi for free data exchange. No cooperative sensing mechanism

or channel allocation mechanism has been proposed.

In [87], a two user model has been discussed in which one acts as a relay sending the information

regarding presence or absence of a PU in the spectrum to other users. Two cooperative schemes

have been discussed. In the first scheme both users detect the PU and first one to detect tells the

other about the presence or absence of a PU signal through a central controller. In the second

scheme, users follow Amplify and Forward (AF) protocol to reduce the detection time. Since

VANET is a multiuser scheme, and each user may have different ideas and findings related to

presence or absence of PU signals, so this proposed scheme is not suitable for a VANET

structure. In [18], the authors have highlighted the challenges and research directions in dynamic

routing emphasizing that a SU (secondary user) suffers from high interference from primary

users and other secondary users. Therefore the secondary transmission must be confined to

ensure sufficient operation of PUs, which thereby deteriorates the QoS provisioning. Due to

channel switching and users’ interference, the longer paths will be chosen for transmissions

which may create a problem of larger routing graphs and end-to-end delay.

In [88], the author has checked the performance of Vehicular Dynamic Spectrum Access

(VDSA) within a TV whitespace environment. A technique for Vehicle Cooperative

Communication and Vehicle Platooning has also been proposed. A test-bed implementing the

VSDA for performance evaluation has also been proposed. In [89], the author has mainly

focused on the spectrum sensing techniques, routing methodology, and security for cognitive

radio vehicular networks. Further, the impact of changes in the network formed by vehicles on

radio propagation channel has been discussed and finally its performance has been evaluated.

In [13] combination of standalone and cooperative spectrum sensing mechanism has been

proposed. It combines the best of both approaches. In this approach, the spectrum is divided into

several non-overlapping channels with a channel spacing of 6, 7, and 8 MHz and the secondary

network has a common channel for the exchange of control information over DSRC (5.9 GHz)

band. A coordinating node (vehicle) periodically senses the spectrum and sends a group of

channels (holes) to the requesting nodes (vehicles). The requesting node on receipt of holes,

rescans the obtained holes and after reconfirmation uses the hole as per class of demand. This

24

approach has its merits and demerits. The merit is that rescanning of holes by requesting nodes

confirms the availability of channel at that time. The demerit is that every channel in a wideband

cognitive radio spectrum may not be a hole for every vehicle due to its dynamic nature and

geographical location.

2.6.3 DETECTION TECHNIQUES AND FADING MODELS

Many researchers have put lot of efforts to analyze the sensing techniques used and fading

models used for CR networks. There are a few works related specifically to the use of CR in

VANETs scenarios. Some existing works are discussed here.

In [85] it is proposed to use energy detection technique over a Gamma-shadowed Nakagami-

composite fading channel [90] for both large and small scale fading in VANETs. Authors have

compared the results obtained in simulation with other composite fading models like Suzuki, Loo

and Rice-Lognormal [91]. In this paper, the authors have not discussed the coordination among

vehicles for sensing the CR network. Researchers in [92] have proposed a model based on

swarm-intelligence (the division of labor model in ant colonies), and claim to reduce congestion

problems to the spectrum database. This paper proposes the use of cellular networks to access the

spectrum database and to use IEEE 802.11p channels for inter-vehicular communication for

cooperative sensing. Vehicles, after sensing, send the results to their neighbors. In [93] authors

have pinpointed the drawbacks of the energy detection technique and have proposed the use of

the covariance based technique. According to the authors, the energy detection technique is

sensitive to noise uncertainty [94] [95] [96]. Furthermore as signals are correlated in most

practical applications hence energy detection should not be considered the method of choice.

The authors propose the computation of the sample covariance matrix of the received signal

based on the received signal samples; then to extract test statistics from the sample covariance

matrix. Finally, a decision is made for the presence or absence of the signal is made by

comparing the ratio of two test statistics with a threshold.

25

2.6.4 SPECTRUM MANAGEMENT AND QOS SUPPORT

Many ideas can be found in relation to spectrum management which also supports the quality of

service requirements. Some interesting ideas relating to dynamic spectrum access have been

explored in [88]. Authors have presented Vehicular Dynamic Spectrum Access (VDSA) to deal

with dynamic spectrum allocation problem being faced by VANETs. For spectrum

measurements and machine learning some VDSA techniques have been presented. As a special

case for the intelligent channel selection reinforced learning is used. (Also presented by [97]).

Secondary Users are modelled as clients and RSU as server. Queuing Theory has been used to

analyse the VDSA technique used. Authors in [98] differed from the idea presented in [88]

pointing out that this technique is over optimistic as vehicles can move out of the range of RSUs.

Hence they have presented an analysis relating to the vehicular dynamics.

Works presented in [84] [75] [99]propose that channel selection algorithm shall allocate the

channel on the basis on QoS requirement of the requesting vehicle according to the specific

location of the vehicle. Another work relating to energy efficiency has been proposed in [100]

presenting an optimization problem whether to use relay base transmission or to use direct

transmission in an objective to minimize energy consumption considering delay as Qos

constraint. These presented approaches are based on centralized architecture. Approach

presented in [81] involves the clustering strategy. It is hybrid of centralised and distributed

approaches. It involves a cluster head and other vehicles acting as cluster members in the group.

In this paper the problem of optimum channel access has been studied in detail in order to

provide QoS support for data transmission. Vehicles use one radio for the DSRC channel and

another for opportunistic spectrum access. For decision making Markov Decision Process has

been used. The cluster head takes the decision if any vehicle will be added in the cluster or not

and also reserves some bandwidth for that vehicle in the DSRC channel. In this way the size of

the cluster is also controlled. Cluster head also collects the sensing information from cluster

members and broadcasts the decision of assigning holes to cluster members. Hence the problem

relating to the opportunistic spectrum access, cluster size control and reservation of some

bandwidth in DSRC channel is solved using linear programming.

26

Some proposals on distributed spectrum management focus on the design of MAC [101] as well

as QoS requirements which may be modified for vehicular environment [102]. Proposal in [102]

insists on one common control channel (CCC) and transmits a beacon frame at regular intervals.

Further it sub-divides beacon in three phases: channel sensing, channel contention and data

transmission. Nodes sense the channel at the first instance, and then use CCC for reservation of

time slot for data transmission. Hence in data transmission phase contention free transmission is

performed by reserving the channel first. QoS is provided by first allowing the priority users for

channel reservation then if slots are available they are utilised for remaining nodes.

2.6.5 DISTANCE SEGMENTATION

In [73] authors have proposed a cooperative sensing Cog-V2V proposal in which every vehicle

collects the sensing data and shares with the vehicles in its range on common CCH channel.

Roads are divided into segments and spectrum horizon of each vehicle is defined. Every vehicle

based on the received information and its own observations for the related road segment and

frequency, aggregates the data and finds the similarity function. It decides the availability of any

channel based on the aggregated data and the similarity function. In [84], the authors have

further enhanced the Cog-V2V framework by considering the impact of correlated shadowing on

the sensing output. Based on correlation, weight factor has been calculated and multiplied with

the sensing binary output to get the decision of primary user’s presence or absence. These two

papers do not take into account the temporal variations in the availability of any spectrum band

in deciding the spectrum band as a hole.

2.7 FUZZY LOGIC

Fuzzy Logic (FL) provides a simpler method to arrive at a definite conclusion based upon vague,

ambiguous, imprecise, noisy, or missing input information. Real world problems are not discrete

(0 or 1). Just like sound may be very low, low, medium, high, very high, and exceptionally high.

FL converts real world problems into membership functions describing graphically the variable

states. Based on those states decisions are made by if-then rules including all possible events and

27

decision is taken based on those rules. Due to several unique features FL is considered a good

choice for many control problems. FL is robust and hence does not need noise free and precise

inputs. Output is a smooth function even in the case of variations in inputs or a wide range of

inputs. FL can be modified easily as it depends on user defined rules. Any number of inputs can

be processed by FL and generating any number of outputs. But to avoid complexity FL should be

broken up into small processes. Nonlinear systems can easily be modelled by a FL system [103].

In [104], the authors have proposed a new evolutionary approach based on a hybrid fuzzy-

particle swarm optimization (PSO) hybrid algorithm to solve the problem of the distribution

networks reconfiguration.

In [105], the authors have designed a Fuzzy-Logic based spectrum allocation algorithm, by

which the RSUs check actual CCH contention conditions, and extend dynamically the CCH

bandwidth in the case of network congestion, by utilizing the detected vacant frequencies by the

sensing module. In [106], the authors have used distance, velocity, signal strength, spectrum

efficiency as input variables to fuzzy logic. Mamdani type fuzzy rule base system has been used.

Based on the fuzzy if-then rules decision is taken for the availability of the spectrum. In [107],

the Spectrum Utilization Efficiency, Degree of Mobility and Distance of Secondary user to the

PU are used as input variables. A fuzzy logic rule base is used to avoid collision among various

contenders for spectrum access. According to authors SU with higher possibility can access the

spectrum with guarantee.

The work in [108] proposes FUZZBR a fuzzy logic based multi-hop broadcast protocol for data

dissemination in VANETS. Authors insist that FUZZBR has low overhead as it uses a subset of

neighbour vehicular nodes for relaying of data. Fuzzy logic has been used to select the relay

node considering different parameters like inter vehicle distance, node mobility and signal

strength. Also here a lightweight retransmission mechanism has been proposed to transmit a

packet in case a relay node fails. Membership functions of distance between vehicles, speed, and

signal strength have been defined to select the relay for which the output function rank defines

the best relay.

28

An algorithm for CR channel selection using fuzzy logic has been proposed in [109]. Idle time

statistics are derived based on historic data collected for the channel occupancy by PUs. In order

to learn the competition level for the available radio channels among secondary users

information exchange based method is used. Then using fuzzy logic, idle time statistics and

competition level among secondary users are integrated into hybrid decision criteria. Longest

value referred by the hybrid decision criteria indicates the most available channel.

29

Chapter 3

STRUCTURE OF VANETS

Structure of VANETs plays an important role in timely dissemination of messages. If the

structure is too complicated, then messages may suffer delays. In case message is urgent, then

delay factor may cause accidents. Moreover, the structure needs to be designed keeping security

in mind. Some of the models have been proposed in literature, like [110] [111]. So the structure

shall be simple and robust in order to reduce the propagation time and increase the security

respectively. Messages also need to be categorized to discriminate while routing immediate,

urgent and ordinary messages.

3.1 MESSAGE CATEGORIZATION

Categorization of messages is important in the sense that these may be treated differently by the

channel according to their priority, authentication requirement, or privacy requirement. This

categorization may save the bulk overhead in the size of messages, improve speed and decrease

occupancy of channel by one message for long time. These messages are categorized as under

[112]:

3.1.1 EMERGENCY MESSAGE

Examples are accident information, congestion information, bridge broken, and train crossing

etc. This type of message needs authentication, but no privacy.

30

3.1.2 SAFETY MESSAGE

Examples are inter-vehicle distance, speed, intersection collision avoidance, and location

information. This type of message also needs authentication, but no privacy.

3.1.3 GPS MESSAGE

This type of message mostly provides road map with reference to location of the vehicle. This

type of message also needs authentication, but no privacy.

3.1.4 PROBE MESSAGE

This is the periodic message for keep-alive between the road side unit and the vehicle. This type

of message needs neither authentication nor privacy.

3.1.5 TRAVELLER INFORMATION

Examples are signal status, road signs, school ahead, hospital ahead, and service ahead etc. This

type of message needs authentication but does not need privacy.

3.1.6 LOCATION BASED SERVICE

Examples are toll collection, and online payments etc. This type of message needs both

authentication and privacy.

3.1.7 INFORMATIVE MESSAGE

The messages involving ordinary internet browsing are informative type of messages. This type

of message needs neither authentication nor privacy.

31

3.1.8 E-MAILS

E-mails usually need both authentication and privacy.

3.2 PROPOSED VANET STRUCTURE

Earlier research on VANETs has not proposed any structure which may cater to the needs of

both urban (city) and Highway/Motorway environments, or the different problems related to

both. We propose a VANET structure as depicted in Figure 3-1. The proposed structure includes

the LCA (Legal Certification Authority) which is the overall controlling authority of the

country's VANET communication. LCA has sub-offices CAs (Certification Authorities) for

certification in different regions to reduce the burden on one office. These issue certificates to

RSUs (Road Side Units), SPs (Service Providers), and the vehicles belonging to individual

regions.

The structure also includes the VCA (Verification and Controlling Authority) having sub-offices

for different regions. VCAs are responsible for the verification of certificates in the individual

regions. In case a certificate is found to be illegal, VCAs have the authority to shut off the

communication of the concerned vehicle or service provider and add them in the RL (Revocation

List). VCAs can also act in case any certificate holder is violating the rules.

The structure includes the SPs which provide different services like toll collection, internet

services, entertainment services (games, audio, video), and location based services like location

of restaurant, fuel station, or workshops. Concrete security measures are required for the services

involving online payments.

32

VCA

VCA 1 VCA 2 VCA 3 VCA m...

LCA

CA 1 CA 2 CA 3 CA m...

SP 1

SP 2

SP n

.

.

.

DRSU DRSU

DRSU

RCU RCU

RSU

Figure 3-1: Overall Proposed Structure of VANETs

The structure includes highways/motorways as well as urban (city) environments. On highways

instead of RSUs, DRSUs (Directional Road Side Units) have been provided. This helps in group

formation on highways. A group of vehicles moving in one direction is easier to manage on

highways as the vehicles moving in one direction stay in the group for a longer period of time

whereas vehicles moving in opposite direction stay in communication with each other for a

shorter duration of time; hence the overhead in the message specially for group formation will be

useless. Cluster formation based on bidirectional traffic as shown in Figure 3-2 is of no use as the

vehicles moving in opposite direction live in cluster for the fraction of a second. We have

proposed cluster formation for unidirectional traffic as shown in Figure 3-3. Fast and Slow

moving vehicles leave the cluster early. Usually vehicles on highways move with specified

normal speed and these vehicles live in clusters for longer duration. Also computational cost

incurred on cluster modification and reformation will be reduced to a greater extent. Hence the

group formation of the vehicles moving in opposite directions is not recommended.

33

Figure 3-2: Cluster Formation for Bidirectional Traffic

Figure 3-3: Cluster Formation for Unidirectional Traffic

In the urban environment vehicles do not stay longer near to each other so the group formation

on the basis of direction is not possible. Instead groups are formed on the basis of categorization

of vehicles into buses, taxies, official vehicles, private vehicles registered by one regional office.

Hence in cities simple RSUs have been provided. It is also proposed to have RCUs (Road

Central Units) instead of RSUs on junctions, with more resources than RSUs to cater for the

communication needs of vehicles moving on different roads joining the junction. This may help

to reduce the infrastructure and installation costs.

3.2.1 SIMULATION RESULTS

The simulation was built using NS2 for highway scenarios. The highway patch is 2 km long,

and the traffic is moving in both directions. CASE-1 is the simulation in which four RSUs are

provided and are forwarding the traffic of both directions whereas CASE-2 is the simulation in

which four DRSUs have been provided. IEEE 802.11 has been used as the MAC layer protocol

34

whereas AODV has been used as the routing protocol. 2.4 GHz band has been used with 22 MHz

channel bandwidth and a total of 14 channels are provided. For adjacent RSUs or DRSUs non

overlapping channels are selected. Radio propagation model used is two ray ground and Omni

antenna has been used for the simulation. Maximum size of queue used for packets is 50 and

droptail type queue is used. In order to simulate real time traffic Constant Bit Rate over User

Datagram Protocol (CBR over UDP) type packet traffic is used. Bit rate for CBR traffic is 64

Kb/s, packet size is 1000 bits and window size is 20 packets. Simulation time is 250 seconds. In

both cases, the simulation has been performed for 20 mobile nodes (10 in each direction) and 50

nodes (25 in each direction). Average throughput, average end to end delay and packet delivery

ratio have been calculated for all cases.

Average Throughput for 20 mobile nodes increased from CASE-1 to CASE-2 by 3.72%, while

for 50 mobile nodes it increased by 47.44%. For CASE-1 with 50 mobile nodes Figure 3-4

shows a plot of throughput vs time. Throughput in this case goes to zero after about 225 seconds

because nodes reach the final destination and get out of range of the RSU.

Figure 3-4: Throughput (CASE-1, 50 Mobile nodes).

For CASE-2 with 50 mobile nodes throughput is shown by the plot in Figure 3-5 below. Spikes

at the end are because some of the nodes have reached the final destination and are out of the

range of DRSU.

35

Figure 3-5: Throughput (CASE-2, 50 Mobile nodes).

Average End to end delay for 20 mobile nodes decreased by 7.26%, whereas for 50 mobile nodes

it decreased by 23.8%.

Packet delivery ratio for 20 mobile nodes increased by 2.93%, whereas for 50 mobile nodes it is

increased by 5.37%. Results in tabular form which shows the percent change in Average

Throughput, End-to-End Delay and Packet Delivery Ratio are given in Table 3.1 below.

Table 3.1: Comparison 20 to 50 nodes, RSU to DRSU

No. of

Nodes

Throughput

Increase

End-to-End

Delay

Decrease

Packet Delivery Ratio

Increase

20 3.72% 7.26% 2.93%

50 47.44% 23.8% 5.37%

3.3 PROPOSED HIGHWAY STRUCTURE

We propose a VANET structure as depicted in Figure 3-6. VCA has the authority to shut off all

communications of the concerned vehicle or service provider and add them in the RL

(Revocation List). VCA can also act in case any certificate holder is violating the law.

36

Figure 3-6: Proposed Highways Structure for VANETs

On highways, instead of RSUs, DRSUs (Directional Road Side Units) have been provided. This

helps in group formation on highways. A group of vehicles moving in one direction is easier to

manage on highways because vehicles moving in one direction stay in a group for a longer

period of time whereas vehicles moving in opposite direction stay in communication with each

other over a shorter duration of time. So, group formation for vehicles moving in different

directions is impractical.


Simulation parameters remain the same as discussed in section 3.2.1. In both cases, the

simulation has been performed for 50 mobile nodes (25 in each direction) and 100 nodes (50 in

each direction). Average end to end delay and packet drop has been calculated for all cases.

VCA

LCA

SP

DRSU DRSU

DRSU

37

Two graphs of throughput (in kbit/s) versus time (in seconds), both concerning 100 mobile

nodes, are given below. Figure 3-7 shows the case where RSUs deal with bi-directional traffic

and Figure 3-8 shows the case where DRSUs deal with unidirectional traffic.

Figure 3-7: Throughput (RSUs catering bi-directional traffic).

Figure 3-8: Throughput (DRSUs catering unidirectional traffic).

Average end to end delay for 50 mobile nodes decreased by 23.8% whereas for 100 mobile

nodes it decreased by 29.3%.

Packets dropped for 50 mobile nodes decreased by 28.3% whereas for 100 mobile nodes they

decreased by 13.3%. We will also notice an increase in the packet delivery ratio. Results in

tabular form are given in Table 3.2 below.

38

Table 3.2: Comparison 50 to 100 nodes, RSU to DRSU

# of

Mobile

Nodes

Decrease in end-

to-end Delay

Decrease in

Dropped Packets

Packet Delivery

Ratio

Increase

50 23.8% 28.3% 5.37%

100 29.3% 13.3% 12.2%

3.4 SUMMARY

In this chapter first of all categorization of the messages is provided in terms of need of

authentication and privacy [4] [112]. Then we proposed a structure for VANETs for highways

based on Directional Road side Units (DRSUs) catering for unidirectional vehicular traffic and

for urban areas based on Road Central Units at crossings and simple RSUs at other locations.

Simulation tests were conducted for DRSUs and found increase in throughput and packet

delivery ratio and decrease in end to end delay and dropped packet ratio.

39

Chapter 4

COGNITIVE RADIO AND VANETS

4.1 PROPOSED SPECTRUM SENSING FRAMEWORK

Here under we propose a spectrum sensing framework based on three coordinators; Main

coordinator, Forward Coordinator, and Backward coordinator. This technique uses a

combination of both cooperative and standalone sensing methods.

4.1.1 NETWORK MODEL

Each vehicle shall be equipped with a CR-Radio capable of sensing and utilizing the available

resources (holes) in the spectrum, a GPS system to track its position and navigate its path to the

intended destination, in addition to the equipment utilizing DSRC IEEE 802.11p resources for

preferred normal communication among the vehicles and between vehicles and RSUs at 5.9 GHz

band. In case DSRC resources fall short of the requirements, vehicles can use CR-spectrum for

their communication.

In USA TV bands use 7-1002 MHz spectrum with VHF low band using 5.9-88 MHz, VHF high

band using 175-216 MHz, and UHF using 470-890 MHz with the channel bandwidth of 6 MHz

CATV uses 7-48 and 55-1002 MHz with 4.5 MHz channel bandwidth. Hence if vehicle uses CR-

spectrum, it may divide the spectrum into k non overlapping channels {C|Ci, i=1, 2, 3, ------, k},

not necessarily equally spaced (in order to utilize the available spectrum resources of various

technologies.) centered at{ }

. Secondary network of vehicles use channel 172 (High

Availability Low Latency Channel) on 5.9 GHz band DSRC spectrum, in order to communicate

40

information and management messages for the use of CR spectrum among vehicles and between

vehicles and RSUs. This channel is referred to as CH.

4.1.2 SPECTRUM SENSING MODEL

The CR network has primary users, which are licensed and must not face any interference from

secondary users. Hence spectrum sensing shall be carried out in such a way to avoid any

interference to primary users’ communication. Every wireless technology has its own pilot

carriers for its different channels. Let indicate the pilot frequency of primary user networks (

indicates the number of channel and * indicate the technology used; such as v for VHF, U for

UHF, and t for CATV). Parallel sensing is used for VHF, UHF, and CATV channels to reduce

the time of sensing the channels. The energy detection technique is used to effectively identify

the presence and absence of primary user signals. Binary hypothesis test can be applied where H1

is used for presence and H0 is used for the absence of primary user signals. As proposed in [113]

and [13] using these two hypothesis conditions, the band pass signal observed by a secondary

user for the channel Ci can be represented as

( ) { {[ ( ) ( )]

}

{ ( )

} (4.1)

Where real part of the complex waveform is represented by Re{.}, i=1,2,3,---,M. M is the

number of CR channels, is the carrier frequency of primary channel, if we use the secondary

user pilot carrier frequency for the detection of secondary user signal occupying the primary

network we use , ( ) is the equivalent low pass representation of the detected primary or

secondary user signal. Additive white Gaussian noise with zero mean is represented by ( ).

Using the energy detection technique and bandpass filter, the energy of detected signal ( ) for

the period can be represented as

∫

( )

(4.2)

41

is a random variable and it has chi-square distribution. Its probability density function can be

expressed as [113] and [13].

( ) {

⁄ ( ⁄ ) ( ⁄ )

⁄

(

⁄ ) (

) ⁄

( ⁄ ) (√ )

(4.3)

Here is the Gamma function and defined as ( ) ∫

, I is the modified Bessel

function, is the instantaneous signal to noise ratio (SNR), and is the degree of freedom.

Situations may arise due to fading if the detected signals have low SNR confusing it with the

noise. In this case caused by misdetection meaning signal is present but it may be considered as

absent. In this case interference may be caused by the secondary users with the communication

of the primary user. There may also be a case of false alarm meaning the signal is absent but is

considered present. But this case causes no harm as far as interference is considered. User

requests are considered random to the main coordinator. Main coordinator furnishes the requests

on the first come first serve basis.

4.1.3 VEHICLE MOBILITY MODEL

Mobility model used for the vehicle mobility is mainly Gipps model [114] with a slight

modification that if the distance between the current vehicle and the leading vehicle is less than

safe distance, then the current vehicle shall overtake the leading vehicle. Other parameters follow

the Gipps model as given below.

( ) ( ) ( ) ( )

( )

(4.4)

( ) [ ( ) ( )] (4.5)

( ) [ [ ( ) ( )]] (4.6)

42

In equations (4), (5), and (6) above, ( ) is the initial speed, ( ) is the speed of leading vehicle

at time t, ( ) is the gap of leading vehicle at time t, is the average speed, ( ) is the

deceleration function, is the driver’s reaction time (usually 1 second), is the maximum

allowable speed of the vehicle, is the acceleration, ( ) is the desired speed, is the

human error factor between 0 and 1, and ( ) is the final speed at time .

4.1.4 SPECTRUM SENSING AND COORDINATION FRAMEWORK

Different ideas have been proposed for the sensing of the CR spectrum. Some are standalone,

where sensing results are faster but they cause unnecessary interference to the primary user

signals and also among secondary users using CR spectrum. Some approaches are cooperation

based which create a master/slave relationship among coordinating node or RSU and secondary

users; also sensing results may not be accurate. Further scalability and intractability problems are

also noticed in these types of approaches. One approach proposed in [13] claim to combine best

of both approaches but the results sensed by a coordinator may not be accurate in high mobility

VANET environment. Our proposed coordinating sensing idea has more than one localized

coordinating node and sensing results are based on a majority decision, which causes more

accuracy, scalability, intractability, and reduces interference.

In the proposed coordination sensing framework, a main coordinator coordinates the sensing

activities and forwards the sensing results to the requesting secondary users based on its sensing

results and the results received from , forward edge coordinator, and , backward edge

coordinator. Sensing results received by from and are finalized using best two out of

three decisions. For example if a channel sensed is considered available (hole) by two or more

coordinators, it is stored as a hole in the database. Requesting secondary users after receiving a

group of channels (holes) from re-sense the channels, but for a slightly longer time compared

to coordinators, confirming these as holes, and then pick a channel suitable for the class of

service requested. Secondary users receive information based on three differently located

coordinators which can be considered more accurate thereby causing negligible interference to

primary user networks.

43

The proposed sensing coordination framework works as follows. The main coordinator and

the two edge coordinators and sense periodically the CR spectrum channels . and

send the sensed results to periodically on channel of DSRC spectrum. Sensing is

performed using the energy detection technique based on hypothesis as given in (4.1). , after

receiving the sensed results makes a decision based on majority for declaring a channel as

occupied or empty (hole) and stores the calculated results in the database. Upon arrival of request

(received from a secondary user on channel of a particular class of service, the

main coordinator sends the sensed results of that class of service to . The requesting

secondary user (after receiving the results), re-senses the received channels, picks a channel

and sends an ACK (acknowledgement) message mentioning the number of channel picked for

use. after receiving an ACK, deletes the channel picked by the from its available channels

database. If the requesting secondary user after re-sensing the channels finds the channels

occupied, it sends a NACK (Non-acknowledgement) message to the , after receipt of which

sends the freshly sensed results to .

a. Coordinators Selection Phase

Coordinators selection is performed dynamically whenever a secondary user wants to access the

CR spectrum. In case of RSU based system, a RSU will act as the coordinator for its coverage

range hence RSUs shall be installed in such a way to fully utilize the CR resources. In case when

RSUs are not available, separate set of coordinators shall be selected for the both sides of

highways. This will ensure that the groups will remain intact for a longer period of time on

highways.

When a secondary user wants to use the CR network for transmission; it firstly checks if there

is any main coordinator available in its vicinity. It confirms this confirming if it is periodically

receiving messages from any . If no main coordinator is available, the requesting node itself

starts behaving as main coordinator . After being selected as , it sends message on

channel to all nodes (vehicles) in its coverage range to send the GPS coordinates and current

speed. When every node replies with the GPS coordinates and speed, selects farthest node in

front with the speed equal or less than (within 5%) the speed of as the front edge coordinator

44

; and farthest node in back with the speed equal to or greater than (with in 5%) the speed of

, as the backward edge coordinator . collects the GPS coordinates and speed of the

cluster (moving cell) nodes periodically and reassign the duty of and to the nodes

matching the criteria as discussed above as due to difference in speed coordinating nodes may be

at different locations than expected.

b. Spectrum Sensing Phase

CR-spectrum sensing responsibility is on the coordinators , and . Each coordinator will

sense the channels { } in parallel, periodically, and in a proactive manner. Each coordinator

will sense the channels independently of each other. Each coordinator shall be able to detect the

presence or absence of the primary or secondary user signals while remaining efficient to reduce

the likelihood of interference. At each iteration of energy detection on channels { } , the

and will send the results on channel to and which will compile the results in terms of

likelihood of channel availability { ( )} according to the following rule.

“The channel will be declared available if and only if at least two of the coordinators will

decide the availability of the channel on the basis of hypothesis (signal absence) or (signal

presence)”.

The probability of detecting the user (primary or secondary) is defined as:

[ ]

[

] (4.7)

In order to obtain , we need to express or depending on whether the PU is present or

absent respectively in the place of , and also we need to derive joint pdf as expressed

later in this section. It is important that during the sensing phase all the coordinators shall be

synchronized. i.e they all shall sense the channel at the same time.

45

At , and the received bandpass waveform on as per equation (4.1) is given by:

( ) {

{[ ( ) ( )]

}

{ ( )

} (4.8)

Here X represents M for main coordinator, F for front coordinator and B for backward

coordinator. As per equation (4.2) the output of energy detection at each coordinator is given by

equation (4.9). In this step it is assumed that mean of noise component is zero and variance is

.

∫

( )

(4.9)

The pdf of individual ( ) is given by equation (4.10).

( )

{

⁄ ( ⁄ ) ( ⁄ )

⁄

(

⁄ )

(

) ⁄

( ⁄ ) (√

)

(4.10)

Since all the coordinators perform the energy detection independently, hence the joint pdf and

third order pdf are given by equations (4.11) to (4.14).

(

) ( ) (

) (4.11)

(

) ( ) (

) (4.12)

(

) ( ) (

) (4.13)

(

) (

) ( ) (

) (4.14)

Therefore probability of detection at can be evaluated as:

46

[

]

[

]

[

]

2 [

] (4.15)

∬ (

)

∬ (

)

∬ (

)

∬ (

)

(4.16)

∫ ( )

∫ ( )

∫ ( )

∫ ( )

∫ ( )

∫ ( )

2∫ ( )

∫ (

)

∫ (

)

(4.17)

47

The detection threshold can be obtained from equation 6 in [13]. Average detection probability

can be obtained by averaging out of over SNR of different signals on CR spectrum as

given by (4.18) below.

∫

⁄

(4.18)

Where is the instantaneous SNR of different (VHF, UHF, CATV) signals on CR spectrum

and is the average SNR.

Upon arrival of the requests from the vehicles, main coordinator send the list of detected

holes to the requesting vehicles on first come first serve basis. Requesting vehicle upon receipt of

list of holes re-senses the received holes multiple times to re-confirm the availability of the

channels and picks the first available channels for the transmission. It sends the channel

acquiring message to the so that it may delete the channel from the list of available channels.

Re-sensing by the requesting vehicle reduces the chances of misdetection and hence avoids

interference with the primary user signals.

4.1.5 COORDINATORS SELECTION AND SENSING ALGORITHMS

Coordinators selection algorithm is given below:

1:

2:

(Front Coordinator selection)

3:

4: (p front nodes)

5: ( ) ( )

48

6:

7: end

(Back coordinator selection)

8:

9: (q back nodes)

10: ( ) ( )

11:

12: end

Algorithm for sensing of the spectrum by three coordinators and compiling the results is given

below:

1: ( )

2:

3:

4:

5:

6:

7:

8: ( ( ) ( ) )

49

( ( ) ( ) )

( ( ) ( ) )

9: ( )

10:

11: ( )

12: end

Flow charts of every activity are given below in figures 4-1 to 4-4.

50

Figure 4-1: Coordinators Selection

Start

Did Nj detect NM?

No

Set NM = Nj

Perform Front Coordinator

Selection

Perform Back Coordinator

Selection

End

Yes End

51

Figure 4-2: Front Coordinator Selection

Figure 4-3: Back Coordinator Selection

Start

Set NF = N1

Set i = 2

i > p? NoD(Ni to Nj) > D(NF to Nj)

Increment iSet NF = NiYes

No

Yes

End

52

Figure 4-4: Spectrum Sensing

4.1.6 SIMULATION AND RESULTS

Simulations have been built using Microsoft C# 5.0. Simulations have been carried out by

varying number of channels, number of vehicles and by changing vehicle speed. The results were

evaluated over a single run, but varied insignificantly over multiple runs. Different scenarios

have been developed for standalone sensing case, cooperative sensing case, and our proposed

spectrum sensing case. A 2 km piece of highway is considered for the simulation, cognitive radio

towers are placed at a perpendicular distance of 91 km from the highway and transmission power

taken is 100 mW. Channels are varied from 10 to 100, number of vehicles from 10 to 100, and

speed from 40 to 110 km/h. Inter vehicle communication range is 240 m and sensing range is

400 m, deceleration factor is 2.12 s-2

, which is , a constant in ( ̅) ̅, and maximum

53

acceleration fraction is 0.01, which is the fraction of average speed a car can accelerate in a

second, Noise power density considered is . This was far too high to be

realistic but was kept high to compare performance of systems under extreme conditions. Power

density threshold is , sensing interval taken is 20 ms, vehicle communication

interval is 20 to 30 ms and vehicle silent time is taken as 0.4 to 0.5 s.

a. Probability of Correct detection

Correctness of the detection technique to identify the presence or absence of primary user signal

is an important factor in determining the validity of the algorithm and robustness of the

equipment used. We have plotted the probability of correct detection versus number of channels,

number of vehicles and vehicle velocity as shown in figure 4-5, figure 4-6, and figure 4-7

respectively. Figure 4-5 shows that as we increase the number of channels on the primary user

network, the probability of correct detection of CR network in the case of standalone and

cooperative sensing methods decreases but in the case of our proposed scheme it remains

constant and almost equal to one. Figure 4-6 shows that if we increase the number of vehicles

probability of correct detection improves with standalone sensing and it decreases with

cooperative sensing (due to load of traffic on a single coordinator and coordination time

required) whereas with our proposed scheme it remains constant at almost one. Figure 4-7 shows

that if we increase the vehicle velocity probability of correct detection with standalone and

cooperative sensing techniques is low as compared to our proposed scheme. From these three

graphs it is evident that our proposed sensing technique outclasses the standalone sensing and

cooperative sensing techniques.

54

Figure 4-5: Probability of Correct Detection versus No. of Channels

Figure 4-6: Probability of Correct Detection versus No. of Vehicles

0.997

0.9975

0.998

0.9985

0.999

0.9995

1

1.0005

0 20 40 60 80 100 120

Co

rrec

t D

etec

tio

n R

ate

# of Channels

Independent Sensing

Proposed Sensing

Cooperative Sensing

0.997

0.9975

0.998

0.9985

0.999

0.9995

1

1.0005

0 20 40 60 80 100 120

Co

rrec

t D

etec

tio

n R

ate

# of Vehicles

Independent Sensing

Proposed Sensing

Cooperative Sensing

55

Figure 4-7: Probability of Correct Detection versus Vehicle Velocity

b. Probability of Misdetection

If any vehicle wrongly identifies the presence of primary user as absence, it can start using that

channel and cause interference with the primary user signals. Hence algorithms and techniques

shall correctly detect the presence of primary user so that secondary user may not harm the

primary user network. Graphs given in figure 4-8, figure 4-9, and figure 4-10 show the

probability of misdetection versus number of channels, number of vehicles, and vehicle velocity.

These graphs show that misdetection caused by our proposed technique is almost eliminated.

Some misdetection is observed in the case of standalone and cooperative sensing techniques

which may cause interference with the primary user network by the unlicensed secondary users

hence misdetection must be eliminated completely. These graphs show that our proposed scheme

have completely eliminated the misdetection and hence does not interfere with the primary user

network.

0.997

0.9975

0.998

0.9985

0.999

0.9995

1

1.0005

60 70 80 90 100 110 120

Co

rrec

t D

etec

tio

n R

ate

Speed (Km/h)

Independent Sensing

Proposed Sensing

Cooperative Sensing

56

Figure 4-8: Probability of Misdetection versus No. of Channels

Figure 4-9: Probability of Misdetection versus No. of Vehicles

-7

-6

-5

-4

-3

-2

-1

0

0 20 40 60 80 100 120lo

g(M

isd

etec

tio

n R

ate)

# of Channels

Independent Sensing

Proposed Sensing

Cooperative Sensing

-7

-6

-5

-4

-3

-2

-1

0

0 20 40 60 80 100 120

log(

Mis

det

ecti

on

Rat

e)

# of Vehicles

Independent Sensing

Proposed Sensing

Cooperative Sensing

57

Figure 4-10: Probability of Misdetection versus Vehicle Velocity

4.2 MODIFIED SPECTRUM SENSING AND ALLOCATION

MODEL

We first lay out a consistent framework for simulating different scenarios. In this framework, we

will vary one factor and keep others constant to achieve consistency in the comparisons. We start

with a “bird’s eye” view of the entire algorithm, and then examine each process in finer detail,

until we are down to the bare essentials. We do this by describing the mobility, sensing,

coordination and scheduling algorithms individually, as well as when they are needed and what

ties them in with the rest of the process.

The proposed framework starts as any time-based simulation model should, by dividing the

entire time over which the simulation is to be performed into a number of time slices, and

simulating over each of those slices.

-7

-6

-5

-4

-3

-2

-1

0

70 80 90 100 110 120 130 140lo

g(M

isd

etec

tio

n R

ate)

Speed (Km/h)

Independent Sensing

Proposed Sensing

Cooperative Sensing

58

Figure 4-11: The Overall Algorithm

The psuedocode corresponding to the flowchart in Figure 4-11 is given below.

nTimeSlices := runningTime/timeStep

for i := 0 upto nTimeSlices - 1

[Perform vehicle movements]

[Perform sensing/coordination/allocation]

next

The procedure described in Figure 4-11 begins by calculating the number of time steps over

which the simulation is to be performed, and then running repetitive procedures over each time

slice. First, we move the vehicles as desired, and as dictated by the mobility model of choice.

Second, we perform channel sensing, coordination and allocation as required by the different

models used.

From first glance, the flexibility of this approach is immediately visible. We can choose the

mobility model while performing movements and we can also choose what methods to use for

59

the other tasks. From here, the first sub-process is completely defined by the choice of mobility

model to be used. The second one needs to be expanded, and it is shown below.

Figure 4-12: Allocation, coordination and sensing


[Perform PU state changes]

if i*timeStep >= nextCoordinationTime

[Perform coordinator selection]

[Perform coordinated sensing]

end if

[Perform sensing and occupation for applicable vehicles]

60

The process shown in Figure 4-12 begins by performing the state changes for PU. This is defined

as the PU (licensed) occupying or vacating the channel in the given time slice. After this, we

check if coordination needs to be performed. If it does, we perform coordination and store results

with the coordinators in a given range. We then perform sensing and occupation for individual

vehicles that need to transmit over the CR spectrum.

Here, the coordinator selection sub-process is defined completely by the choice of coordination

algorithm used. The other two need to be expanded, as they are below. Note that these processes

are for individual channels or vehicles.

Start

Is the PU scheduled to

occupy or vacate

channel?

Vacate

Tear down transmission channel as scheduled.

Occupy

Schedule next occupation of

channel

End

None

Is there SU on the channel?

Tear down SU communication

Yes

Start transmission.

No

Schedule vacation of channel

Figure 4-13: How to perform PU state changes.


if i*timeStep >= vacationTime && occupying

[Tear down transmission channel]

[Set occupationTime in accordance with scheduling algorithm]

61

occupying := false

else if i*timeStep >= occupationTime && !occupying

if occupied by SU

[Tear down SU transmission]

end if

[Start transmission]

[Set vacationTime in accordance with scheduling algorithm]

occupying := true

end if

Figure 4-13 defines the basic scheduling algorithm used for PU state changes. It defines how the

simulation will decide when a PU is to occupy a channel, and when it is to vacate a given

channel.

Again, a lot of flexibility as to the method of scheduling transmissions is allowed. A message

must be sent to SUs occupying this channel, possibly over the WAVE control channel.

62

Start

Is the car scheduled to

occupy or vacate

channel?

Vacate

Tear down transmission channel as scheduled.

OccupyGet channel(s)

from coordinator (if applicable)

Sense all applicable

channels and attempt to occupy

first

Schedule next occupation of

channel

Was a channel occupied?

No

Back off for a certain time period

YesSchedule next

vacation of channel

End

None

End

Figure 4-14: Sensing and occupation mechanism.


if i*timeStep >= vacationTime && occupying

[Tear down transmission channel]

[Set occupationTime in accordance with scheduling algorithm]

occupying := false

else if i*timeStep >= occupationTime && !occupying

if coordinated sensing scheme

[Get channel(s) from coordinator]

end if

[Attempt to sense/Occupy as necessary]

if occupation was successful

[Set vacationTime in accordance with scheduling

algorithm]

occupying := true

63

else

[Extend occupationTime by back-off time]

end if

end if

These procedures, shown in Figure 4-14, define the (more complicated) scheduling and

allocation scheme for SUs of the CR spectrum. It takes into account the fact that a channel may

not always be allocated to a SU when needed, and it also lays the groundwork for what sensing is

to be performed before a SU will attempt to occupy a channel.

Please note that this last flowchart may vary slightly depending on the needs of the coordination

and allocation scheme. For example, coordination may not be required at all; multiple channels

may be sensed before a back-off is triggered, and so on. The flowchart has been made for the

requirements of [12], but the pseudo-code is more generalized. This also takes into account the

possibility of a misdetection.

4.2.1 MODELS USED FOR THE SIMULATION

a. Hata Model

The Hata Model [19] for suburban areas was used as the large-scale fading model in this

simulation. This depends on the Hata model for urban areas. This gives the predicted median

path loss over a given path. Equation (4.19) describes the median path loss in urban

environments, in decibels.

[ ] (4.19)

Equations (4.20) and (4.21) describe the calculation of in cities with low to medium

populations and cities with a high population, respectively.

( ) (4.20)

64

{ ( )

( )

(4.21)

In equations (4.19), (4.20) and (4.21); is the median path loss for urban areas in decibels,

is the base station antenna height in meters, is the mobile station antenna height, is the

carrier frequency in MHz, is the antenna height correction factor and is the distance in

kilometers.

Equation (22) defines how to calculate the path loss in a suburban area.

(

)

(22)

In Equation (22), is the predicted median path loss for suburban areas in decibels. This

second formula is the one that was used.

The Hata model is a simple model that simulates large scale urban fading. It takes into account

obstacles that may be encountered, and gives the median path loss that a given signal will

experience under given conditions. The simulation for a particular point, which will account for

instantaneous variations in power, is given by Rayleigh fading, which is fed the median power

flux density calculated by the Hata model.

b. Rayleigh Fading

Whereas the Hata model is a large-scale model that predicts averages over a given distance,

Rayleigh fading predicts sudden changes due to rapidly varying channel conditions. Equation

(4.23) gives the probability density function for a Rayleigh random variable.

( )

⁄ (4.23)

In Equation (4.23), is the mode of the distribution. Since the median of this distribution is

given by √ , we can easily use this model for a given median as well, as required by

the Hata model [20].

65

Rayleigh fading is a long-standing model used in wireless communication systems for power

prediction, where a line-of-sight component is absent. Considering the scenario, which is a

suburban environment, it is highly unlikely that one will be present. Therefore, we have chosen

the Rayleigh model for our simulation.

c. Gipps Mobility Model

The Gipps mobility model describes how traffic behaves in a typical vehicular scenario.

Equations (4.4), (4.5) and (4.6) describe how to calculate the new velocity of a vehicle given the

old velocity in a given time step [114].

The Gipps model is one of the classical models used to simulate urban mobility. We choose it for

its simplicity and effectiveness in predicting vehicle behavior. We made one modification to this

model for our purposes: That the vehicle will not go below a certain minimum speed threshold.

Given we are simulating a highway scenario where vehicles may overtake each other, this

assumption is reasonable. Equation (4.24) is the replacement we use for equation (4.6) in our

simulations.

( ) [ [ ( ) ( )]] (4.24)


Here, we give simulation results for three different sensing and allocation schemes, for three

different kinds of data. We vary the number of available channels, the number of vehicles in the

simulation and the average speed of the vehicles. The schemes used are standalone sensing,

cooperative sensing, and proposed sensing, as used in [115]. The other parameters were: Vehicle

sensing range was 400 m and communication range was 240 m, the length of road used was 2

km, the noise power was -133.16 dBW (500 K system noise temperature at a bandwidth of 7

MHz). Two PU transmission towers were used, each 10 km away from the closest point on the

road. The mobility model used was the Gipps model [114] and the wave propagation model used

was the Hata model [19] for large scale fading and the Rayleigh model [20] for small scale

66

fading. The frequency used was 150 MHz, the base station height was 50 m and the mobile user

height was 1.5 m. The average speed for vehicles was set to 100 km/h ± 20% per vehicles (when

not varied), the number of channels to 100 (50 per PU tower) and number of vehicles to 50 (25

per side). A vehicle could deviate to within 10% of its average speed. The reaction time was set

to 1 second, and the deceleration function was 2.12 times velocity. The human error factor could

vary between 0.25 and 0.4 for each vehicle. The vehicle back-off time was 10 ms and the

coordination time was 20 ms.

The recorded data in the simulations:

a) The number of successful allocations of a channel.

b) The number of false alarms (a channel was vacant but was sensed as occupied).

c) The number of misdetections (a channel was occupied but was sensed as vacant).

Misdetections are dangerous because they can interfere with PU or other SU

communications.

The following figures 4-15 to 4-23 show the results of simulations run under a number of

scenarios.

Figure 4-15: Allocations Rate vs Vehicles

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 20 40 60 80 100 120

Allo

cati

on

Rat

e

# of Vehicles

Independent Sensing

Proposed Sensing

Cooperative Sensing

67

Figure 4-16: False Alarms vs Vehicles

Figure 4-17: Misdetections vs Vehicles

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 20 40 60 80 100 120lo

g(Fa

lse

Ala

rm R

ate)

# of Vehicles

Independent Sensing

Proposed Sensing

Cooperative Sensing

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 20 40 60 80 100 120

log(

Mis

det

ecti

on

Rat

e)

# of Vehicles

Independent Sensing

Proposed Sensing

Cooperative Sensing

68

Figure 4-18: Allocations Rate vs Channels

Figure 4-19: False Alarms vs Channels

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0 20 40 60 80 100 120

Allo

cati

on

Rat

e

# of Channels

Independent Sensing

Proposed Sensing

Cooperative Sensing

-4.5

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 20 40 60 80 100 120

log(

Fals

e A

larm

Rat

e)

# of Channels

Independent Sensing

Proposed Sensing

Cooperative Sensing

69

Figure 4-20: Misdetections vs Channels

Figure 4-21: Allocations Rate vs Speed

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

0 20 40 60 80 100 120lo

g(M

isd

etec

tio

n R

ate)

# of Channels

Independent Sensing

Proposed Sensing

Cooperative Sensing

0

0.05

0.1

0.15

0.2

0.25

0.3

70 80 90 100 110 120 130 140

Allo

cati

on

Rat

e

Speed (Km/h)

Independent Sensing

Proposed Sensing

Cooperative Sensing

70

Figure 4-22: False Alarms vs Speed

Figure 4-23: Misdetections vs Speed

The differences between the proposed model and the other schemes were:

a) Cooperative sensing had cluster formation on both sides, and just one coordinator per

cluster. Every vehicle senses the spectrum and sends the result to coordinator. In this

-3

-2.5

-2

-1.5

-1

-0.5

0

70 80 90 100 110 120 130 140lo

g(Fa

lse

Ala

rm R

ate)

Speed (Km/h)

Independent Sensing

Proposed Sensing

Cooperative Sensing

-4

-3.5

-3

-2.5

-2

-1.5

-1

-0.5

0

70 80 90 100 110 120 130 140

log(

Mis

det

ecti

on

Rat

e)

Speed (Km/h)

Independent Sensing

Proposed Sensing

Cooperative Sensing

71

model, the coordinator compiles the result and allocate the channel, but cars blindly

followed the coordinator instead of sensing it themselves before occupation.

b) Independent sensing had no clusters at all and each car sensed the channel independently

before occupying it.

Independent sensing had a lack of coordination, due to which errors were high in most cases. In

cooperative sensing, clusters were plagued by vehicles leaving and entering the clusters at high

rates, because of which coordination was rendered difficult.

In our model, cluster formation was concrete and a three coordinator system established high

certainty of the result of sensing the spectrum. The number of allocations was slightly lower due

to a small delay coming from coordination, but it was offset by a smaller number of false alarms

and (more importantly) misdetections.

As one can clearly see, in all three cases, the number of misdetections is almost eliminated. In

this limited simulation, they actually came out to be nearing zero in all cases. This is shown in

Figure 4-17, 4-20 and 4-23. The number of false alarms is also greatly reduced, as shown in

Figure 4-16, 4-19 and 4-22. One can see a fall in the allocation rate as compared to standalone

sensing and cooperative sensing in Figure 4-15, 4-18 and 4-21. This is because coordination may

have an adverse effect on the allocations, due to the time taken to communicate channels to and

from the coordinator. Further as the false alarms are greatly reduced and misdetections are

almost eliminated the reduction in allocation rate is eminent. In any case, the benefits outweigh

this slight disadvantage, particularly that of the elimination of misdetections. It is very important

that a SU should not interfere with a PU’s communication. This is achieved in the proposed

sensing model. Graphs show the standard deviation and error bars not visible on certain graphs

are too small to be displayed.

4.3 SUMMARY

In this chapter we have proposed a cognitive radio sensing model for VANETs based on three

coordinators (Main, Front, and Back). Sensing is performed by the three coordinators and results

72

are prepared on best two out of three basis. Gibbs Mobility model has been used for vehicle

mobility. In simulations probability of correct detection and probability of misdetection have

been calculated and compared with standalone and cooperative approaches and found

considerable improvement. Then we used Hata model for large scale fading, Rayleigh Fading

model for small scale fading and suggested a modification in Gibbs Mobility model. Simulation

results show a considerable improvement in allocation rate and decrease in false alarms and

misdetections.

73

Chapter 5

HISTORY BASED SPECTRUM

SENSING AND ALLOCATION

Here, we propose a system which relates to suburban areas and highways. It is understood that RSUs are

installed at proper distances on the sides of highways. For efficient and systematic communication among

vehicles and between vehicles and infrastructure Directional RSUs (DRSUs) shall be installed as

proposed in Chapter 3. As DSRUs have separate processors and storage locations for each direction,

cluster management can be achieved easily because clusters remain intact for longer times on separate

directions of any highway. Vehicles are managed in clusters on each direction and each cluster has three

coordinators: (Main Coordinator), (Front Coordinator), and (Back Coordinator) for CR

spectrum sensing and coordination activities as proposed in section 4.1. The Cognitive Radio spectrum,

especially the TV spectrum, is divided into M channels. Communication range of is divided into

fifteen equal distance segments denoted by [73], [84] as shown in Figure 5-1. is moving in distance

segment , in , and in ; which are progressively changing as vehicles move forward and

jump into the next distance segment. is the third last segment of the backward cluster and carries the

front coordinator of that cluster. is the third segment of front cluster and carries the back

coordinator of that cluster. The segments and are overlapping segments between the

backward and current cluster; whereas the segment and are overlapping segments between the

front and current cluster. It is important to mention here that highway considered here is busy, but as

traffic conditions and patterns change dynamically and may be some of the segments contain no vehicles

74

(in shallow traffic conditions) then positions of back and front coordinators may shift one or two

segments front or backwards. The selection of the coordinators was defined in section 4.1.

Figure 5-1: Segment distribution in a cluster

Every vehicle sends and receives the control information on the control channel (CCH) of 5.9 GHz

IEEE 802.11p/DSRC spectrum. Vehicles in segments to send and receive control messages to

; vehicles in segments to to ; and vehicles in segments to to . Vehicles in

segments , also send and receive control messages to and vehicles in segments ,

also send and receive control messages to to receive the information on next segments from front

cluster or to provide information to the backward cluster. Further every vehicle is equipped with a GPS

(Global Positioning System) chip in order to locate its position and track the roads and intended

destination.

5.1 HISTORY UPDATING

Every vehicle (Secondary User) senses the M channels of TV spectrum every second using the

energy detection technique in order to identify the presence or absence of licensed users (Primary Users)

in the spectrum. Binary hypothesis test is conducted; indicating for the presence of primary user and

for the absence of primary user as given by equation 1.

Every after sensing sends a sense message containing the results (Binary 1 for and binary 0 for )

for every channel it sensed along with its coordinates indicating the distance segment and also its identity

75

in order to avoid duplication of messages on CCH channel to , , or depending on its location in

the cluster. Each entry in the list contained in the message is formatted as follows.

< X Y CH RES Time>

Where is the vehicle identity, X and Y are the coordinates received from GPS to locate the segment,

CH is the channel sensed from 1 to M, RES is the sensing result (1 or 0), and Time is the sensing time. It

shall be noted that in to and to also send the sense message to and

respectively in order to be aware of the future locations on the path. Figure 5-2 shows the cluster

formation in a RSU based network. Instead of RSUs we have provided Directional RSUs (DRSUs) as

proposed in Chapter 3.

Figure 5-2: Clusters in a RSU based Network

, , or on receipt of sensed messages from the vehicles for one interval compile the results

based on majority decision if more than one result is obtained from one segment for the same channel

f. Results for the space, time, and frequency are finalized and sent to the main coordinator for onward

submission to the RSU. A sample entry in the list contained in message sent to RSU is of the format as

given below. Output indicates the binary 1 for primary or secondary user presence and binary 0 for

absence. RSUs on receipt of update message updates in spatial-time-frequency database.

76

< f Output>

5.2 HISTORY PRESERVATION

RSU divides the twenty four hours of the day into K time slots. Each time slot is further subdivided

into S time intervals. For each data entries are collected from the clusters for every distance segment

and every channel. One time slot containing S time intervals is represented in Figure 5-3.

Figure 5-3: Contents of Tk time slot

Probability for the availability of the channel is computed by the RSU based on number of zeros in the

time slot (5.1) and a database for K time slots, each channel, and each segment is stored. Database entries

computed and stored by the RSUs will look like as shown in Figure 5-4.

(5.1)

Figure 5-4: Database maintained by the RSUs

77

RSUs exchange the database with each other and each RSU stores the database for the segments covered

by its basic service set (BSS) and also for the segments covered by the BSS of next RSU. This will enable

the vehicles to utilize the spectrum proactively on their future path to the destination.

5.3 HISTORY UTILIZATION

Each RSU sorts the database in order of high probability for the next time slots and shares the information

with main coordinators of the clusters for the segments covered by that main coordinator and its future

path. The main coordinator shares the information with and . All three coordinators sense the

channels as per procedure discussed in [115], compile the results and send back to main coordinator. ,

on receipt of results from itself and the other two coordinators, compiles the results based on current

availability and high availability probability received from RSU for the segments in its coverage range.

On receipt of CR spectrum utilization request from any moving in segment and time slot ,

sends back the channel numbers in order of higher probability of availability to the lower one for the next

distance segments and time slot . Sample format of reply sent by to the requesting will look like

as shown in Table 5-1.

Table 5-1: reply to

Segment Channels

12 5 3 14 8

7 5 1 2 9

13 11 3 7 4

3 12 15 8 6

78

on receipt of reply from the as per format in Table 5-1, rescans the channels in order of higher to

lower probability and utilizes the first available channel for its communication and continuously scans the

channels for the next segments and switches to other channels if required on its way to the destination.

5.4 SIMULATION RESULTS

Simulations have been performed using Visual C#. Spectrum sensing mechanism used is as presented in

section 4.2 which is the modified version of section 4.1. The mechanism for the movement of vehicles as

well as other parameters has been lifted from sections 4.1 and 4.2. The only thing added on is spatio-

temporal history, databases and predictive sensing as discussed below. As shown in Figure 5-5, whenever

a vehicle wants to use the CR-spectrum, it checks the sensing data it has available in its own database,

corresponding to the current time and space segment. If unavailable, it requests new history data for its

time segment and future segments, and similarly for space, from the coordinators.

79

Figure 5-5: Data Collection and Sensing

Similarly, if the coordinator finds that its own data are inadequate, it requests data from the RSU

in batches, for its entire communication range and then some more, so it won’t have to request

new data too much.

Sensing is performed as detailed in [21], but with a minor difference: Channels with a higher

probability of being empty (also for a longer amount of time) in the given spatio-temporal slot

are sensed first; with the first available channel being occupied.

80

This process is then reversed, with simple (non-coordinating) vehicles sending data to the

coordinators and the coordinators sending it to the RSUs, with the RSUs updating daily records.

Figures 5-6 and 5-7 detail the process.

Figure 5-6: Data Calculation

81

Figure 5-7: History Records Updating

There were a number of different parameters used during the simulation, which are listed as

follows: the vehicle communication range was 240 meters, the average car speed varied from 80

km/h to 120 km/h, with each individual car varying its speed between 5% to 10%. Noise power

was watts (measured by the popular formula for Additive White Gaussian

Noise , where B is the bandwidth (7 MHz for a typical TV channel), k is the Boltzmann

constant and T is the noise temperature (we used 500 Kelvin). The threshold used to measure if

another user was present was watts. The Hata model for suburban areas was used for

fading, with height of transmitting antenna = 50 m, height of receiving antenna = 1.5 m, and the

center frequency = 1.5 GHz. Cars would communicate for 20 to 30 ms, and stay “silent” for 0.4

to 0.5 seconds. The patch of road was 2 km long.

A number of different metrics have been observed in these simulations. The most important of these

metrics are Allocation Rate, False Alarms, Misdetections, Rejection Rate, and Forced Leaves. Allocation

82

Rate is the ratio of number of successful allocations to the number of total attempts hence it provides a

clear picture of the success of the algorithm. False Alarm indicates an empty channel detected busy and

the ratio is number of false alarms to the number of total attempts. Lower false alarms are better for

efficient resource utilization but are not harmful for the spectrum and primary user. Misdetection

indicates a busy channel detected empty. Misdetections are very harmful as it leads to interference with

the primary user signals and are needed to be eliminated. Our algorithm as discussed in section 4.1 and

section 4.2 has successfully eliminated Misdetections. Rejection Rate tells the number of un-successful

attempts to the total number of attempts. Lower Rejection Rate is a measure of success of the algorithm.

Forced Leave ratio indicates the number of times once the vehicles acquire the channels is forced to leave

the channels due to the return of primary user. Lower Forced Leave ratio indicates the fewer disturbances

to the communication of the secondary users.

These metrics have been measured in a number of different scenarios: varying numbers of cars, number

of CR-spectrum channels and varying average speed of the vehicles, keeping all other factors constant.

Results observed for Allocation Rate are given by graphs in Figure 5-8, Figure 5-9 and Figure 5-10. It is

clear from the graphs that Allocation Rate for our proposed Algorithm is better as compared to the

Independent sensing and Cooperative sensing mechanisms. In Figure 5-8, a higher number of cars cause

congestion in the spectrum, resulting in a lower allocation rate. In Figure 5-9, the proposed model has

lower results in a low number of channels due to it being heavily reliant on PU timings. The lower the

channels, the lower the reliance on the PU timings. In Figure 5-10, it was noted that the number of

requests were higher when there were drops in the graph, resulting in a lower allocation rate. We may see

an increase in the allocation rate compared to the technique presented in section 4.2. Results obtained for

False Alarms Ratio are given by Figure 5-11, Figure 5-12 and Figure 5-13. Results show that False Alarm

ratio of our proposed algorithm is better than Independent sensing mechanism but Cooperative sensing is

better in this respect. This is due to time slots being large: it was possible that in the time slot, the PU had

83

vacated the channel, but it wasn’t added to the empty channels list due to a low probability in the rest of

the time slot. As False Alarms are not harmful so it is not a problem if we obtain Allocation Rate higher,

Rejection Rate lower and Forced Leave Ratio lower. Figure 5-14, Figure 5-15 and Figure 5-16 show the

Rejection Rate graphs. It is evident from the graphs that our proposed algorithm has a lower rejection rate

compared to the other two mechanisms. Figure 5-17, Figure 5-18 and Figure 5-19 indicate the Forced

Leave ratio. The forced Leave Ratio of our proposed mechanism is zero. This is due to the fact that our

proposed algorithm monitors the channels and calculates the higher probability of availability for the

duration of the intended duration of the transmission. Hence it provides the list of those channels which

will be available for a longer duration calculated based on the history. All the graphs show the error bars

specifying standard deviation. Error bars not visible on some graphs are too small to be displayed.

Figure 5-8: Allocation Rate Vs No. of Cars

84

Figure 5-9: Allocation Rate Vs No. of Channels

Figure 5-10: Allocation Rate Vs Speed

85

Figure 5-11: False Alarm Rate Vs No. of Cars

Figure 5-12: False Alarm Rate Vs No. of Channels

86

Figure 5-13: False Alarm Rate Vs Speed

Figure 5-14: Rejection Rate Vs No. of Cars

87

Figure 5-15: Rejection Rate Vs No. of Channels

Figure 5-16: Rejection Rate Vs Speed

88

Figure 5-17: Forced Leave Ratio Vs No. of Cars

Figure 5-18: Forced Leave Ratio Vs No. of Channels

-6

-5

-4

-3

-2

-1

0

10 20 30 40 50 60 70 80 90 100Lo

g(Fo

rce

d L

eav

e R

ate

)

No. of Cars

Proposed Sensing

Independent Sensing

Cooperative Sensing

-6

-5

-4

-3

-2

-1

0

10 20 30 40 50 60 70 80 90 100

Log(

Forc

ed

Le

ave

Rat

e)

No. of Channels

Proposed Sensing

Independent Sensing

Cooperative Sensing

89

Figure 5-19: Forced Leave Ratio Vs Speed

5.5 SUMMARY

In this chapter we have proposed to build a database based on spatial-temporal frequency slots

for the road segments in different time and different frequency channels. Results are obtained

from the moving vehicles and stored in a database at DRSUs and three coordinators discussed

earlier and are continuously updated giving a proper weightage to history and current results.

Allocation is performed based on better probability of availability of channels during that

particular time and for that road segment. Results obtained are compared with other popular

approaches and have found considerable improvement.

-6

-5

-4

-3

-2

-1

0

70 80 90 100 110 120 130 140Lo

g(Fo

rce

d L

eav

e R

ate

)

Speed Km/h

Proposed Sensing

Independent Sensing

Cooperative Sensing

90

Chapter 6

CR CHANNEL ALLOCATION USING

FUZZY LOGIC IN VANETS

It has been observed that some wireless bands especially TV and radio channels have a specific

on/off criteria in a day. Based on this observation a system can be designed to utilize the

spectrum available due to off timings of these channels. We have modeled this system using

fuzzy logic [116] as detailed in this chapter.

6.1 SYSTEM MODEL

Our fuzzy logic system is multi-input multi-output (MIMO). The main purpose of the system is

the selection of the channels from TV channels available on Cognitive Radio (CR) network

based on different input scenarios. The block diagram of our system is given in Figure 6-1

below.

Figure 6-1: Overall Fuzzy Logic System

91

6.1.1 INPUTS AND MEMBERSHIP FUNCTIONS

Our system is a six input eight output system. The first input is the speed of the vehicle. The set

of linguistic values for speeds are very slow, slow, medium, fast, and very fast. Vehicles moving

specifically on highways have a speed from zero to 150 km/h. This range is however different

from models used in chapters 4 and 5 but we wanted to check the system at higher range. Speed

plays an important role in channel assignment. As vehicles move in clusters and sensing is based

on control by the coordinators of the cluster [115] [117], so the vehicles moving with very fast or

very slow speed leave the cluster early whereas vehicles moving with medium speed stay in the

cluster for a longer duration. Thus, vehicles moving with very fast or very slow speed shall be

allocated those channels which are available for a shorter duration of time and the vehicles

moving with medium speed shall be allocated the channels which are available for a longer

duration of time. The speed of the vehicles is assumed to follow the truncated Gaussian

distribution [13]. The membership function of speed is given in Figure 6-2.

Figure 6-2: Membership Function of Speed

The second input linguistic variable is message priority. The set of linguistic values for message

priority are immediate, urgent, normal and ordinary. Immediate messages include emergency

messages like accident information, congestion information, bridge broken, train crossing, and

safety messages like inter-vehicle distance, speed, intersection collision avoidance, and location

information. Urgent messages include traveler’s information like signal status, road signs, school

ahead, hospital ahead, service area ahead and location based services like toll collection, and

online payments. GPS and probe messages fall in the category of normal messages. While net-

surfing and emails fall in the category of ordinary messages [4]. Message category plays an

92

important role in the channel allocation mechanism. Higher priority messages need to have stable

and longer duration channels, whereas lower priority messages can have channels with lower

stability or duration. The membership function in trapezoid shape is shown in Figure 6-3 below.

Figure 6-3: Membership Function of Message Priority

The third input linguistic variable is time in hours of the day. Linguistic values for the time are

mid-night, morning, noon, after-noon, evening and night. Different TV channels have different

on/off timings. Some channels are on late night, some during evening and some during mid-day.

So time is an important factor for the determination of channel allocation of TV channels. The

membership function in trapezoid shape of time is shown in Figure 6-4.

Figure 6-4: Membership Function of Time

Channel sensing is crucial in the channel allocation scheme. We use the spectrum sensing results

obtained by using techniques proposed in our previous work [115] [117] [118]. In this chapter

we have used the sensing result based on the categorization of the different types of TV channels

93

as discussed in next paragraph. We have combined the sensing results of channels 1 to 3 in one

input variable, 4 to 6 in another input variable and 7 to 8 in another input variable.

6.1.2 OUTPUTS

Output variables are eight different types of channels. We have categorized these channels as;

channel-1 is entertainment channel, channel-2 is news channel, channel-3 is infotainment

channel, channel-4 is kids channel, channel-5 is informative channel, channel-6 is educational

channel, channel-7 is food channel and channel-8 is midnight programs channel. These different

categorized channels have different on/off timings. We have formed the membership function for

the availability (off timing) of each type of channel as shown in Figures 6-5 to 6-12.

Figure 6-5: Availability of Channel-1



95




96



We have taken on/off timings of above categorized channels as arbitrary as different countries

have different timings for TV channels. Anyone willing to use this technique has to alter the

timings according to their own available timing slots.

6.1.3 FUZZY IF-THEN RULES

By using vehicle speed, message priority, hours of the day and channel sensing results we can

model the fuzzy IF-THEN rules based the criteria discussed in above paragraphs. Our system

presented in this paper has been modeled using 320 rules. Some rules are given below for better

understanding.

97

If (Speed is Very-Fast) and (Message-Priority is Normal) and (Time is Noon) then (CH1

is not T1)(CH2 is not T8)(CH3 is not T12)(CH4 is T18)(CH5 is not T21)(CH6 is not

T24)(CH7 is not T29)

If (Speed is Very-Slow) and (Message-Priority is Immediate) and (Time is Noon) then

(CH1 is not T1)(CH2 is not T6)(CH3 is not T12)(CH5 is not T21)(CH6 is not T24)(CH8

is T35)

If (Speed is Fast) and (Message-Priority is Ordinary) and (Time is Evening) and (Ch-1-3-

Sensing-Results is not T9) and (Ch-4-6-Sensing-Results is T26) then (CH1 is not

T3)(CH2 is not T9)(CH3 is not T14)(CH4 is not T18)(CH5 is not T23)(CH6 is T26)(CH7

is not T31)(CH8 is not T35)

If (Speed is Slow) and (Message-Priority is Normal) and (Time is Evening) and (Ch-4-6-

Sensing-Results is T26) then (CH1 is not T3)(CH3 is not T14)(CH4 is not T20)(CH5 is

not T23)(CH6 is T26)(CH7 is not T31)(CH8 is not T35)

If (Speed is Medium) and (Message-Priority is Immediate) and (Time is Night) and (Ch-

7-8-Sensing-Results is T35) then (CH1 is not T3)(CH2 is not T9)(CH3 is not T14)(CH4

is not T20)(CH5 is not T23)(CH7 is not T33)(CH8 is T35)

A widely used implication in fuzzy systems and controls is Mamdani Implications. The argument

in the favor of Mamdani Implications is that it considers fuzzy IF-THEN rules as local. For

global implications other type of implications are available.

6.2 SIMULATION AND RESULTS

Simulations have been carried out using MATLAB version R2012a, Fuzzy Logic Toolbox.

Mamdani Implications have been used to model the Channel Selection Fuzzy model. Input,

output linguistic variables used and membership functions modeled are described in detail in

System Model section above. Three hundred and twenty IF-THEN rules have been used to

model the system. Results have been obtained checking the utility of the proposed system by

98

looking at the utility of the channels in speed-time domain and in message priority-time domain.

These utility results actually show the allocation of that particular channel. Figures 6-13 to 6-20

show the utility of channel-1 to 8 in speed-time domain and Figures 6-21 to 6-28 show the utility

of the same channels in message priority-time domain.

6.2.1 UTILITY SPEED VS TIME

Figure 6-13: Speed Vs Time for Channel 1


99




100




From the above figures it is evident that channel-1 is well utilized for very fast and very slow

speeds. Utility of channel-2 is better in early hours of the day. Channel-3 is best utilized for all

the speeds during its availability timings. Channel-4 is better for slow and fast speeds in the early

hours of the day. Channel-5 is better for all speeds during morning hours. Channel-6 is better for

very slow and very fast speeds in night hours. Channel-7 is better in early hours for very slow,

101

slow, fast and very fast speeds. Channel-8 is better during available hours for very slow, very

fast and medium speeds but its utility for slow and fast speeds drops during noon and after-noon

hours. Utility results indicate the allocation of these channels for different speed and time.

6.2.2 UTILITY MESSAGE PRIORITY VS TIME

Figure 6-21: Message Priority Vs Time for Channel 1


102


Figure 6-24:Message Priority Vs Time for Channel 4


103




From the above figures it is clear that channel-1 is well utilized during morning hours for all

priority messages. Channel-2 is better during first half of the day for all priorities except

immediate. Channel-3 is better in early hours for all priorities and in late hours for immediate

and urgent messages. Channel-4 is better for lower priority messages and channel-5 for all

104

priorities during morning hours. Channel-6 is better for higher priority messages during late

hours. Channel-7 is better during morning hours for all sort of messages and channel-8 very well

utilized especially for immediate and urgent messages. These results indicate the allocations for

different message priority and time.

6.3 SUMMARY

In this chapter we have proposed a fuzzy logic based cognitive radio channel allocation scheme

for the vehicles. Channel sensing model used is based on as proposed in chapter 5. These

channels sensing results along with speed, time in hours of the day and message priority are

taken as inputs and allocation possibilities are obtained. Better allocation possibilities are

achieved by using this model as shown by results.

105

Chapter 7

CONCLUSION AND FUTURE

PROSPECTS

7.1 CONCLUSION

As the vehicles move with high speed on highways they switch from one RSU to another very

quickly. Hence it will be difficult for bidirectional Road Side Units to bear the load of the traffic

nodes on both sides. Furthermore as on highways speed of most of the vehicles remains in the

similar range with the minor difference of 5% to 10 % hence the formation of groups for

management and security purposes is easier for the vehicles moving in one direction. But with

bidirectional RSUs it will be difficult to form the groups/clusters for vehicles moving in both

directions. With the introduction of directional Road Side Units and from simulation results it is

observed that average throughput is increased considerably when the traffic on the road increases.

Also the end to end delay is reduced by a better margin with the increase in traffic. Packet

delivery ratio is also increased. Also the groups/cluster formation will be easier and manageable.

As discussed in the introduction section many different techniques have been proposed for

spectrum sensing of cognitive radio networks. Since primary users are licensed users and are

authorized to utilize that network hence their communication shall not be disturbed in any case

by the unlicensed secondary users. So the technique and algorithm used shall correctly identify

the presence or absence of primary user and it shall force the secondary users to leave the

primary network channel in case primary user comes back again. As proposed in section 4.1 and

shown by the simulation results our technique has a better probability of correct detection as

compared to standalone and cooperative sensing techniques further our technique has almost

106

eliminated misdetection resulting in negligible interference being caused with the primary user

signals. Since vehicular communication is needed to be error free and quick so our technique

provides the opportunity to utilize CR networks in case the resources on the DSRC spectrum are

overburdened.

In section 4.2, we introduced a framework for the impartial comparison of CR-VANET sensing

and allocation schemes. We described the framework and all the models that could be used to fill

in various parts of the algorithm. We then described the schemes we used to produce a complete

simulation. We then ran simulations in line with the framework for three different channel

sensing and coordination schemes. The proposed coordination scheme with three coordinators

per cluster showed marked benefits over independent and cooperative sensing schemes.

Day by day the vehicular traffic is increasing on the roads; so it is the need of the hour to search

for the vacant frequency slots in CR-spectrum and then utilize until these are occupied again by

the primary users. Our proposed algorithm discussed in section 4.3 is based on the historic data

of sensing the CR-spectrum so it provides a clear picture of spatio-temporal and frequency slots

for its future activity. It observes the primary user’s activity and timing related to acquiring and

leaving the channel for a particular distance slot and frequency. Based on these computations and

observations a list is prepared giving priority to the channels which are most likely to be

available for the duration of the intended transmission. The simulation results obviously show

the success of our algorithm.

The model proposed in chapter 5 provides a clear picture while allocation of CR channels to the

moving vehicles depending upon speed, message priority, hours of the day, and sensing results.

Considering the utility of channels, priority in allocation can be achieved easily. There may be

more than one channels of one category available at a time with same instants availability or

different. This model can be utilized well for all sorts of channels depending on sensing results.

107

7.2 FUTURE PROSPECTS

In future it is proposed that opposite side RSUs may be integrated into one RSU to reduce the

cost of the hardware to be installed. Further work may also be carried out to check formation of

groups for localized traffic and security issues. Further work may also be carried out on

formation of groups categorized in buses, trams, taxies, cars, vans etc. for urban traffic

environments and improvements in security. In urban areas Road Central Units (RCUs) may be

installed at crossings and simulations may be performed. Vehicular traffic in urban areas is a

major concern for researchers. Hence a concrete structure is needed for growing need of

VANETs in urban areas.

Day by day traffic in urban areas as well as on highways is increasing leaving researchers to

think about providing sufficient resources to cope with the increasing vehicle to vehicle and

vehicle to infrastructure communication. Our technique may help to distribute the burden of

increased communication and avoid unwanted incidents.

As the concept of the application of CRs is relatively new to VANETs, there is a lot of room for

improvement. Further work needs to be done on the application of effective spectrum

management and allocation in VANETs. This is particularly true for CR-VANETs, as there is a

possibility of PU involvement. Furthermore, work needs to be done on the possibility of using

bandwidth-efficient techniques. As power generation isn’t a huge issue, work needs to be done

on immersive entertainment systems while inside a vehicle, while improving the safety. Vehicle

to vehicle communication in real-time may open up the door for maintaining a safe distance from

other vehicles, with facilities for auto-braking in the case of risks. Other similar features for lane

guidance and optimal speed may also help improve the utility of vehicles in general.

We are also working to add channel bandwidth, fading, end to end delay, packet delivery ratio,

packet drop ratio and distance segments as further input variables in the model proposed in

chapter 6 to enhance the capability of the proposed system. This picture will provide a realistic

environment in vehicular networks and will help to build a stable system for VANETs.

108

Appendix A

TOOLS USED

During my research work I have used following different tools for the simulations of my

proposed models.

1 NS2

2 C#

Some discussion on these tools is given below.

NS2

Network Simulator version 2 is open source event driven simulation tool specially designed for

simulating different scenarios in computer communication networks. NS2 was started in 1989

and since then various researchers from education, industry, and government sectors have used it

for their valuable proposals. It is continuously under enhancements and contains modules for

application, transport, network and data link layer simulations. Various network performance

parameters can be integrated in NS2 in order to measure the objects of researchers using easy to

use scripting language. There may be some modules needed by researchers which are beyond the

scope of NS2 but such modules can be integrated in NS2 by programmers using back end object

oriented language. For this purpose profound knowledge of NS2 architecture is required. Some

of the widely used modules are Network AniMator (NAM) for visualizing animations and

Xgraph for generating graphs of the results. As the researchers are developing new modules for

their specific purposes; NS2 developers are integrating new modules enhancing its performance.

Due to its flexibility and modular built up NS2 has gained too much popularity. University of

109

California and Cornell University have played an important role in the development of real NS2

simulator. Virtual InterNetwork Testbed (VINT) project with the support of Defense Advanced

Research Projects Agency (DARPA) was main cause for the development of Network Simulator

[119].

NS2 consists of mainly two key languages. C++ and OTcl (Object Tool Command Language).

OTcl sets up simulation by the assembly and configuration of the objects whereas C++ works as

the back end. Both languages are linked together by TclCL (TCL with classes). The variables in

OTcl objects are called handles which are mapped to C++ objects. Handle acts as a front end for

interaction with users and other OTcl objects. There is one to one correspondence between C++

and OTcl hierarchies. C++ has compiled and OTcl has interpreted Hierarchies. In compiled

hierarchy member or class variables and member or class functions are referred to as variables

and functions respectively. In interpreted hierarchy these are referred to as instance variables and

instant procedures [119].

NS2 is supported by FreeBDS, Linux, Solaris, Windows and Mac. NS2 supports Wired as well

as Wireless networks. In wired networks the features supported are Routing, Transportation,

Traffic sources, Queuing, disciplines, and QoS. In Wireless netowrks Ad hoc routing, mobile IP,

and sensor-MAC are supported. NS2 creates trace file for graphs and data manipulations and for

visualizing NAM file is generated. NS2 supports applications and traffic models such as FTP,

CBR, Telnet, Web and real audio. It supports unicast transport protocols such as TCP and UDP

and multicast such as SRM. It supports RED and Drop-tail queuing protocols [120] [121].

C#

O VISUAL STUDIO

Visual Studio developed by Microsoft is an Integrated Development Environment (IDE) which

can be used to develop console applications and graphic user interface (GUI) applications. It can

also be used to develop window form applications, web pages, web applications in both managed

110

as well as native code for all supported platforms of Microsoft windows, window CE, .NET

framework, windows phone, Microsoft silvernight and .NET compact framework [122].

O C# HISTORY AND FEATURES

C# pronounced as ‘see sharp’ was developed by Microsoft in the year 2000. Its main designer

and lead architect was Anders Hejlsberg. He wanted to design a new programming language

which could create class libraries in .NET framework. It was understood that C# is imitation of

Java. However in spite of strong object oriented approaches both have distinct features. C# has

support for automatic garbage collection; array bound checking, and software engineering

principles. It has software components which can be integrated for development in distributed

environment. It has source code similarity features as in C or C++.

C# understands everything as object. It reflects directly the CLI (Common Language

Infrastructure). Its intrinsic types are similar to value types in CLI framework. It is more type

safe than C++. It strongly supports operator overloading like C++ and unlike Java. Due to

garbage collection property memory leak problems are avoided. C# has try…finally construct in

addition to try…catch construct which guarantees (in case an exception occurs or not) execution

of code in finally block. C# unlike Java does not has checked exceptions. In order to simplify

architectural requirements throughout CLI, it does not support multiple inheritance. Structs are

different in C# than classes. These are not reference types but value types so they are passed by

value. These are not derived from any base class so they are sealed. Unlike C++ inheritance is

always public in C#. A class in C# can only be derived from one base class. In case base class is

not specified then class will automatically be derived from System Object. C# permits delegates,

events and properties in addition to C++ class members such as variables, functions,

constructors, destructors and operator overloads. In C# two more access modifiers are

introduced. Those are a) Internal, b) Protected Internal [123]

111

O .NET FRAMEWORK

This framework primarily runs on Microsoft Windows. It has large library and supports many

programming languages. This feature supports language inter-operability. Its library supports all

the programming languages supported by .NET framework. Programmes written with the

support of .NET framework can execute in a unique software environment called CLR (Common

Language Runtime). Services provided by CLR are security, exception handling and memory

management. So .NET framework is the combination of CLR an application virtual machine and

class libraries [122].

112

BIBLIOGRAPHY

[1] WHO, "Road Traffic Injuries," World Health Organization, 2015. [Online]. Available:

http://www.who.int/mediacentre/factsheets/fs358/en/.

[2] "Global Status Report on Road Safety," World Health Organization, 2013.

[3] "Fata Analysis Reporting System," National Highway Traffic Safety Administration,

[Online]. Available: http://www-fars.nhtsa.dot.gov.

[4] S. H. Abbassi, I. M. Qureshi, O. Khalid and H. Abbasi, "Basic Structural Change in

Vehicular Adhoc Networks," Journal of Networking Technology, vol. 4, no. 3, pp. 149-

154, Dec.2013.

[5] "Sensor and Vehicle Networks," 24 February 2012. [Online]. Available:

http://sensorandvehiclenetworks.blogspot.com/2012/02/primera-toma-de-contacto-

hsvn.html.

[6] L. Figueiredo, I. Jesus, J. Machado, J. Ferreira and J. Martins de Carvalho, "Towards the

development of intelligent transportation systems," in IEEE Intelligent Transportation

Systems, Hong Kong, 2001.

[7] A.-S. K. Pathan, "Security of Self-Organizing Networks," CRC Press, Taylor and Francis

Group, 2011.

[8] S. Haykin, "Fundamental Issues in Cognitive Radio," in Cognitive Wireless

113

Communication Networks, Springer US, 2007, pp. 1-43.

[9] P. D. Barjoei, G. Dadashzadeh, F. Razzazi and S. M. Razavizadeh, "Power and Time Slot

Allocation in Cognitive Relay Networks Using Particle Swarm Optimization," The

Scientific World Journal, vol. 2013, no. 2013, 2013.

[10] L. Giupponi and C. Ibars, "Distributed cooperation in cognitive radio networks: Overlay

versus underlay paradigm," in 69th IEEE Vehicular Technology Conference, Barcelona,

2009.

[11] K. B. Letaief and W. Zhang, "Cooperative Spectrum Sensing, Cognitive Wireless

Communication Networks," Springer Science + Business Media LLC., pp. 115-138, 2007.

[12] J. Chen, X. Zheng and M. Zhu, "Cooperative Spectrum Sensing Algorithm based on the

Relay," Procedia Engineering, vol. 15, no. 2011, pp. 2349-2353, 2011.

[13] X. Y. Wang and P.-H. Ho, "A Novel Sensing Coordination Framework for CR-VANETs,"

IEEE Transaction on Vehicular Technology, vol. 59, no. 4, pp. 1936-1948, 2010.

[14] P. Mankar, S. S. Pathak and R. V. Rajakumar, "A Cooperative Secondary User

Localization based Primary User Localization Method for Cognitive Radio Networks," in

National Conference on Communications, Kharagpur, 2012.

[15] J. Mitola and G. Q. Maguire, "Cognitive radio: Making software radios more personal,"

IEEE Personal Communications, vol. 6, pp. 13-18, 1999.

[16] S. Haykin, "Cognitive radio: brain-empowered wireless communications," IEEE J. Sel.

Areas Commun, vol. 23, pp. 201-220, 2005.

[17] W. Zhang, R. K. Mallik and K. Letaief, "Cooperative Spectrum Sensing Optimization in

Cognitive Radio Networks," in IEEE International Conference on Communications, 2008.

114

[18] S. Tauqeer and B. H. Halabi, "Interference-Aware Routing in Cognitive Radio Ad-hoc

Networks: Chanlenges and Research Directions," MAGNT Research Report, vol. 2, no. 3,

2014.

[19] M. Hata, "Empirical Formula for Propagation Loss in Land Mobile Radio Service," IEEE

Transactions Vehicular Technology, vol. 29, no. 3, pp. 317-325, August 1980.

[20] B. Sklar, "Rayleigh Fading Channels in Mobile Digital Communication Systems Part I:

Characterization," IEEE Communications Magazine, vol. 35, no. 7, pp. 90-100, July 1997.

[21] H. Abbasi, S. H. Abbassi and I. M. Qureshi, "A framework for the simulation of CR-

VANET channel sensing, coordination and allocation," Journal of Chinese Institute of

Engineers, vol. Submitted, 2014.

[22] S. H. Abbassi, I. M. Qureshi, H. Abbasi and B. R. Alyaie, "History Based Spectrum

Sensing in CR-VANETs," vol. 2015:163, 2015.

[23] U. Universitet, "Introduction to Mobile Ad hoc Networks (MANETs)," [Online].

Available: http://user.it.uu.se/~erikn/files/DK2-adhoc.pdf.

[24] M. Zitterbart and K. Weniger, "Mobile ad hoc networks (MANET)," TeleMatics, 2005.

[25] C. Imrich, C. Marco and J.-N. L. Jennifer, "Mobile ad hoc networking: imperatives and

challenges," Adhoc Networks, vol. 1, no. 2003, pp. 13-64, 2003.

[26] A. Tønnesen, "Mobile Ad-Hoc Networks," [Online]. Available: www.olsr.org/docs/wos3-

olsr.pdf.

[27] J. Hoebeke, I. Moerman, B. Dhoedt and P. Demeester, "An Overview of Mobile Ad Hoc

Networks: Applications and Challenges," Journal of the Communications Network, vol. 3,

no. July 2004, pp. 60-66, 2004.

115

[28] Aarti and S. S. Tyagi, "Study of MANET: Characteristics, Challenges, Application and

Security Attacks," International Journal of Advanced Research in Computer Science and

Software Engineering, vol. 3, no. 5, pp. 252-257, 2013.

[29] R. K. Sakib and B. Reza, "Security Issues in VANET," BRAC University, Dhaka,

Bangladesh, 2010.

[30] T. Taleb and E. Sakhaee, "A Stable Routing Protocol to Support ITS Services in VANET

Networks," IEEE Transactions on Vehicular Technology, vol. 56, no. 6, pp. 3337-3347,

2007.

[31] S. Wasiq, W. Arshad, N. Javaid and A. Bibi, "Performance Evaluation of DSDV, OLSR

and DYMO using 802.11 and 802.11p MAC-Protocols," in IEEE 14th International

Multitopic Conference (INMIC), Karachi, 2011.

[32] S. Sagar, J. Saqib, A. Bibi and N. Javaid, "Evaluating and Comparing the Performance of

DYMO and OLSR in MANETs and in VANETs," in IEEE 14th International Multitopic

Conference (INMIC), Karachi, 2011.

[33] P. I. Offor, "Vehicle Ad Hoc Network (VANET): Safety Benefits and Security

Challenges," Social Science Research Network, Nova Southeastern University, 3 Dec.

2012.

[34] D. V. Jamthe and S. S. Dorle, "Collision Avoidance in Ivan to maintain Inter-Vehicular

Distance on Highways," International Journal of Engineering Science and Advanced

Technology, vol. 2, no. 3, pp. 672-678, 2012.

[35] J. Antonio, G. Ibanez, C. Flores-Cortes and J. M. Ramirez-Alcaraz, "Intelligent Signaling

for Prevention of Intersection Collision in Urban Zones," in 8th International Conference

on Intelligent Environment, 2012.

116

[36] Irshadullah and S. u. Rehman, "Analysis of Black Hole Attack on MANETs Using

Different MANET Routing Protocols," School of Computing, Blekinge Institute of

Technology, Sweden, 2010.

[37] Cruz-Cunha and M. Manuela, Handbook of Research on Mobility and Computing:

Evolving Technologies and Ubiquitous Impacts, IGI Global, 2011.

[38] G. Xu, "Location cloaking for location privacy protection and location safety protection,"

Iowa State University, 2010.

[39] M. Rezvani, A. Ignjatovic, E. Bertino and S. Jha, "Secure Data Aggregation Technique for

Wireless Sensor Networks in the Presence of Collusion Attacks," IEEE Transactions on

Dependable and Secure Computing, vol. 12, no. 1, pp. 98-110, 2014.

[40] "DSRC: The Future of Safer Driving," Intelligent Transportation Systems, [Online].

Available: http://www.its.dot.gov/factsheets/dsrc_factsheet.htm.

[41] IEEE, "P802.11p/D5.0, IEEE 802.11 Amendment 7: Wireless Access in Vehicular

Environments," Nov. 2008.

[42] J. B. Kenney, "Dedicated Short-Range Communications (DSRC) Standards in the United

States," Proceedings of the IEEE, vol. 99, no. 7, pp. 1162-1182, July 2011.

[43] E. R. Nyanhete, M. Mzyece and K. Djouani, "Operation and Performance of Cognitive

Radios in Vehicular Environments," Tshwane University of Technology, Aug. 2011.

[44] IEEE, "Std 1609 family, IEEE Trial-Use Standard for Wireless Access in Vehicular

Environments(WAVE)," Nov. 2006.

[45] D. Jiang and D. Luca, "IEEE 802.11p: Towards an International Standard for Wireless

Access in Vehicular Environments," in IEEE Vehicular Technology Conference,

117

Singapore, 2008.

[46] K. Bilstrup, E. Uhlemann and E. G. Ström , "Medium access control in vehicular networks

based on the upcoming IEEE 802.11p standard," in 15th World Congress on Intelligent

Transport Systems, 2008.

[47] J. Kakarla, S. S. Sathya and B. G. Laxmi, "A Survey on Routing Protocols and its Issues in

VANET," International Journal of Computer Applications, vol. 28, no. 4, Aug. 2011.

[48] A. Reyes, C. Barrado and M. López, "Vehicle density in VANET applications," Journal of

Ambient Intelligence and Smart Environments, vol. 6, no. 4, pp. 469-481, 2014.

[49] W. Alasmary and W. Zhuang, "Mobility impact in IEEE 802.11p infrastructureless

vehicular networks," Ad Hoc Networks, vol. 10, no. 2, pp. 222-230, 2012.

[50] N. Loulloudes, G. Pallis and M. D. Dika, "The Dynamics of Vehicular Networks in Urban

Environments," Cornell University Library, 2011.

[51] Y. Toor, P. Muhlethaler and A. Laouiti, "Vehicle ad hoc networks: applications and related

technical issues.," IEEE Commun. Surv. Tutor., vol. 10, no. 3, pp. 74-88, 2008.

[52] H. Hartenstein and K. Laberteaux, "A tutorial survey on vehicular ad hoc networks," IEEE

Communication Magzine, vol. 46, no. 6, pp. 164-171, 2008.

[53] S. Kukliński and G. Wolny , "CARAVAN: Context-AwaRe Architecture for VANET," in

Mobile Ad-Hoc Networks: Applications, Rijeka, Croatia, Intech, 2011.

[54] K. Sha, S. Wang and W. Shi, "RD4: Role-Differentiated Cooperative Deceptive Data

Detection and Filtering in VANETs," IEEE Transactions on Vehicular Technology, vol.

59, no. 3, pp. 1183-1190, 2010.

118

[55] S. Mahajan and A. Jindal, "Security and Privacy in VANET to reduce Authentication

Overhead for Rapid Roaming Networks," International Journal of Computer Applications,

vol. 1, no. 20, pp. 17-21, 2010.

[56] A. Studer, E. Shi, F. Bai and A. Perrig, "TACKing Together Efficient Authentication,

Revocation, and Privacy in VANETs," in 6th Annual IEEE Communications Society

Conference on Sensor, Mesh and Ad Hoc Communications and Networks, Rome, 2009.

[57] A. Frederik, F. Andreas, W. Dirk and Z. Ke, "Cross-layer Privacy Enhancement and Non-

repudiation in Vehicular Communication," in ITG-GI Conference on Communication in

Distributed Systems, Bern, Switzerlad, 2007.

[58] N. Balon and J. Guo, "Increasing broadcast reliability in vehicular ad hoc networks," in

3rd international workshop on Vehicular ad hoc networks, Los Angeles, 2006.

[59] K. A. Hafeez, L. Zhao, B. Ma and J. W. Mark, "Performance Analysis and Enhancement

of the DSRC for VANET’s Safety Applications," IEEE Transactions on Vehicular

Technology, vol. 62, no. 7, pp. 3069-3083, 2013.

[60] B. Hughes, R. Meier, R. Cunningham and V. Cahill, "Towards real-time middleware for

vehicular ad hoc networks," in First ACM International Workshop on Vehicular Ad Hoc

Networks, Philadelphia, 2004.

[61] M. Torrent-Moreno, J. Mittag, P. Santi and H. Hartenstein, "Vehicle-to-vehicle

communication: Fair transmit power control for safety-critical information," IEEE

Transactions on Vehicular Technology, vol. 58, no. 7, pp. 3684-3703, 2009.

[62] E. M. Vaneenennaam, W. Kleinwolterink, G. Karagiannis and G. J. Heijenk, "Exploring

the solution space of beaconing in VANETs," in 1st IEEE Vehicular Networking

Conference, Tokyo, Japan, 2009.

119

[63] K. Bilstrup, E. Uhlemann, E. G. Strom and U. Bilstrup, "Evaluation of the IEEE 802.11p

MAC method for vehicle-to-vehicle communication," in IEEE 68th Vehicular Technology

Conference, Calgary, BC , 2008.

[64] Z. Wang and M. Hassan, "How much of DSRC is available for non-safety use?," in 5th

ACM Int. Workshop on Vehicular Inter-Networking, San Francisco, California, USA,

2008.

[65] M. X. and X. B. Chen, "Delay and broadcast reception rates of highway safety applications

in vehicular ad hoc networks," in Mobile Networking Vehicular Environment, Anchorage,

AK, 2007.

[66] S. Eichler, "Performance evaluation of the IEEE 802.11p WAVE communication

standard," in IEEE Vehicular Technology Conference, Baltimore, MD, 2007.

[67] W. Ejaz, N. U. Hasan and H. S. Kim, "Stand-Alone Architecture for Spectrum Sensing in

Cognitive Radio Networks for Smart Buildings," in 2nd International Conference on

Networking and Information Technology, 2011.

[68] X. Y. Wang, A. Wang and P. H. Ho, "Prioritized spectrum sensing in cognitive radio based

on spatiotemporal statistical fusion," in IEEE Wireless Communications and Networking

Conference, Budapest, 2009.

[69] H. T. Cheng, H. Shan and W. Zhuang, "Infotainment and road safety service support in

vehicular networking: From a communication perspective," Journal of Mechanical

Systems and Signal Processing, vol. 25, no. 6, pp. 2020-2038, 2011.

[70] W. Chen, "Effect of adjacent channel interference on uplink capacity in a macrocell‐

microcell‐overlapping WCDMA system," Journal of the Chinese Institute of Engineers,

vol. 31, no. 1, pp. 157-161, 2008.

120

[71] K. Popper, The Logic of Scientific Discovery, Routledge, 1959.

[72] K. Fawaz and A. Ghandour, "Improving Reliability of Safety Applications in Vehicle Ad

hoc Networks through the Implementation of a Cognitive Network," in 17th International

Conference on Telecommunications, Doha, 2010.

[73] M. Di Felice, K. R. Chowdhury and L. Bononi, "Analyzing the Potential of Cooperative

Cognitive Radio Technology on Inter-Vehicle Communication," in Proceedings of the

2010 IFIP Wireless Days, Venice, 20-22 Oct. 2010.

[74] W. Kim and S. Y. Oh, "CoRoute: A New Cognitive Anypath Vehicular Routing Protocol,"

in Proceedings of the 2011 7th International Wireless Communications and Mobile

Computing Conference, Istanbul, 4-8 July 2011.

[75] L. Husheng and D. K. Irick, "Collaborative Spectrum sensing in Cognitive Radio

Vehicular Ad hoc Networks: Belief Propagation on Highways," in Proceedings of the 2010

71st IEEE Vehicular Technology Conference, Taipei, 16-19 May 2010.

[76] D. B. Rawat, Y. Zhao and G. Yan, "CRAVE: Cognitive Radio Enabled Vehicular

Communications in Heterogeneous Networks," in Proceedings of the 2013 IEEE Radio

and Wireless Symposium, Austin, 20-23 Jan. 2013.

[77] I. Akyildiz, B. Lo and R. Balakrishnan, "Cooperative spectrum sensing in cognitive radio

networks: a survey," Physics Communications, vol. 4, no. 1, pp. 40-62, 2011.

[78] A. Min and K. Shin, "Impact of mobility on spectrum sensing in cognitive radio

networks," in ACMWorkshop on Cognitive Radio Networks, Beiging, China, 2009.

[79] D. Borota, G. Ivkovic, R. Vuyyuru, O. Altintas, I. Seskar and P. Spasojevic, "On the delay

to reliably detect channel availability in cooperative vehicular environments," in IEEE

121

73rd Vehicular Technology Conference, Yokohama, 2011.

[80] R. Doost-Mohammady and K. Chowdhury, "Design of spectrum database assisted

cognitive radio vehicular networks," in 7th International ICST Conference on Cognitive

Radio Oriented Wireless Networks and Communications, Stockholm, 2012.

[81] D. Niyato, E. Hossain and P. Wang, "Optimal channel access management with QoS

support for cognitive vehicular networks," IEEE Transactions on Mobile Computing, vol.

10, no. 4, pp. 573-591, 2011.

[82] I. H. Brahmi, S. Djahel and Y. Ghamri-Doudane, "A Hidden Markov Model based Scheme

for Efficient and Fast Dissemination of Safety Messages in VANETs," in Proceedings of

the 2012 IEEE Global Communication Conference, California, 3-7 Dec. 2012.

[83] A. R. Biswas, T. C. Aysal, S. Kandeepan, D. Kliazovich and R. Piesiewicz, "Cooperative

Shared Spectrum Sensing for Dynamic Cognitive Radio Networks," in Proceedings of the

2009 IEEE International Conference on Communications, Dresden, 14-18 June 2009.

[84] M. Di Felice and K. R. Chowdhury, "Cooperative Spectrum Management in Cognitive

Vehicular Ad Hoc Networks," in Proceedings of the 2011 IEEE Vehicular Networking

Conference, Amsterdam, 14-16 Nov. 2011.

[85] H. Rasheed and N. Rajatheva, "Spectrum Sensing for Cognitive Vehicular Networks over

Composite Fading," International Journal of Vehicular Technology, vol. 2011, pp. 1-9,

2011.

[86] F. Riaz, I. Shafi, S. F. Hasan, W. Younas and Y. Mehmood, "Vehicle-to-Vehicle

Communication Enhanced by Cognitive Approach and Multi-Radio Technologies," in

Proceedings of the 2012 IEEE International Conference on Emerging Technologies,

Islamabad, 8-9 Oct. 2012.

122

[87] G. Ganesan and L. Ye, "Cooperative Spectrum Sensing in Cognitive Radio, Part I: Two

User Networks," IEEE Transactions on Wireless Communications, vol. 6, no. 6, pp. 2204-

2213, 2006.

[88] S. Chen, "Vehicular Dynamic Spectrum Access: Using Cognitive Radio for Automobile

Networks," Worcester Polytechnic Institute, 2012.

[89] K. Komathy, "Application of Cognitive Radio in VANET," Anna University, India.

[90] I. Kostic, "Composite Phase PDF in Gamma Shadowed Nakagami Fading Channel,"

Wireless Personal Communications, vol. 41, no. 4, p. 465, 2007.

[91] C. Press, Signal Processing for Mobile Communications, Taylor & Francis Group, 2004.

[92] M. Di Felice, A. J. Ghandour, H. A. Hartail and L. Bononi, "Integrating Spectrum

Database and Cooperative Sensing for Cognitive Vehicular Networks," in 78th IEEE

Vehicular Technology Conference, Las Vegas, Nevada, USA, 2-5 Sept. 2013.

[93] Y. Zeng and Y.-C. Liang, "Spectrum-Sensing Algorithms for Cognitive Radio Based on

Statistical Covariances," IEEE Transactions on Vehicular Technology, vol. 58, no. 4, pp.

1804-1815, May 2009.

[94] A. Sonnenschein and P. M. Fishman, "Radiometric Detection of Spread-Spectrum Signals

in Noise of Uncertainty Power," IEEE Trans. Aerosp. Electron. Syst., vol. 28, no. 3, pp.

654-660, July 1992.

[95] R. Tandra and A. Sahai, "Fundamental Limits on Detection in Low SNR Under Noise

Uncertainty," in International Conference on Wireless Networks, Communications and

Mobile Computing, Wuhan, 13-16 June 2005.

[96] D. Cabric, A. Tkachenko and R. W. Brodersen, "Spectrum Sensing Measurements of Pilot,

123

Energy, and Collaborative Detection," in IEEE Military Communications Conference.

MILCOM 2006, Washington DC, 23-25 Oct. 2006.

[97] S. Chen, R. Vuyyuru, O. Altintas and A. Wyglinski, "On optimizing vehicular dynamic

spectrum access networks: automation and learning in mobile wireless environments," in

IEEE Vehicular Networking Conference, Amsterdam, 2011.

[98] M. Khabbaz, C. Assi and A. Ghrayeb, "Modeling and analysis of DSA-based vehicle-to-

infrastructure communication systems," IEEE Transactions on Intelligent Transport

Systems, vol. 14, no. 3, pp. 1186-1196, 2013.

[99] I. F. Akyildiz, W. Y. Lee, M. C. Vuran and S. Mohanty, "A Survey on Spectrum

Management in Cognitive Radio Networks," IEEE Communications Magazine, vol. 46, no.

4, pp. 40-48, April 2008.

[100] C. Yang, Y. Fu, Y. Zhang, S. Xie and R. Yu, "Energy-efficient hybrid spectrum access

scheme in cognitive vehicular ad hoc networks," IEEE Communications Letters, vol. 32,

no. 18, pp. 1983-1997, 2009.

[101] S. Jha, M. Rashid, V. Bhargava and C. Despins, "Medium access control in distributed

cognitive radio networks," IEEE Transactions on Wireless Communications, vol. 18, no. 4,

pp. 41-45, 2011.

[102] S. Jha, U. Phuyal, M. Rashid and V. Bhargava, "Design of OMC-MAC: an opportunistic

multi-channel MAC with QoS provisioning for distributed cognitive radio networks,"

IEEE Transactions on Wireless Communications, vol. 10, no. 10, pp. 3414-3425, 2011.

[103] S. D. Kaehler, "Fuzzy Logic Tutorial," Encoder. [Online].

[104] R. Mohadeseh, T. Hamid and R. Anise, "Reconfiguration Optimization for Loss Reduction

in Distribution Networks Using Hybrid PSO Algorithm and Fuzzy Logic," MAGNT

124

Research Report, vol. 2, no. 5, pp. 903-911, 2014.

[105] A. J. Ghandour, K. Fawaz, H. Artail, M. D. Felice and L. Bononi, "Improving vehicular

safety message delivery through the implementation of a cognitive vehicular network,"

Journal of Adhoc Networks, vol. 11, no. 8, pp. 2408-2422, Nov. 2013.

[106] P. P. Bhattacharya, R. Khandelwal, R. Gera and A. Agarwal, "Smart Radio Spectrum

Management for Cognitive Radio," International Journal of Distributed and Parallel

Systems, vol. 2, no. 4, pp. 12-24, july 2011.

[107] K. R, D. N. Kumar C. and P. A., "Fuzzy Logic System for Opportunistic Spectrum Access

using Cognitive Radios," International Journal of Computer Science, vol. 10, no. 1, pp.

703-709, Jan. 2013.

[108] C. Wu, S. Ohzahata and T. Kato, "VANET Broadcast Protocol based on Fuzzy Logic and

Lightweight Retransmission Mechanism," IEEE Transactions on Communications, Vols.

E95-B, no. 2, pp. 415-425, 2012.

[109] Y. Yong, P. Alexandra and P. Adrian, "On Fuzzy Logic Based Channel Selection in

Cognitive Radio Networks," in Handbook of Research on Software-Defined and Cognitive

Radio Technologies for Dynamic Spectrum Management, University of North Dakota,

USA, Information Science Reference, 2014, pp. 255-274.

[110] S. Khan and N. Chauhan, "Region Authority (RA) Collaborated Certificate Organization

and Management in VANET," in World Congress in Computer Science Computer

Engineering and Applied Computing, Las Vegas, Nevada, USA, 2014.

[111] L. Mohan, "Security in VANETs," 1 December 2014. [Online]. Available:

http://www.cse.wustl.edu/~jain/cse571-14/index.html .

125

[112] J. SAE, "Dedicated Short Range Communications (DSRC) message set dictionary".

[113] F. F. Digham, M. S. Alouini and M. K. Simon, "On the energy detection of unknown

signals over fading channels," IEEE Transactions on Communication, vol. 55, no. 1, pp.

21-24, 2007.

[114] P. G. Gipps, "A Behavioral Car Following Model for Computer Simulation," Journal of

Transportation Research, vol. 15, no. 2, pp. 105-111, 1981.

[115] S. H. Abbassi, I. M. Qureshi, B. R. Alyaei, H. Abbasi and K. Sultan, "An Efficient

Spectrum Sensing Mechanism for CR-VANETs," Journal of Basic and Applied Scientific

Research, vol. 12, no. 3, pp. 365-378, 2013.

[116] S. H. Abbassi, I. M. Qureshi, H. Abbasi and B. R. Alyaie, "CR Channel Allocation Using

Fuzzy Logic in VANETs," MAGNT Research Report (ISSN. 1444-8939), vol. 3, no. 3, pp.

898-909, 2015.

[117] H. Abbasi, S. H. Abbassi and I. M. Qureshi, "A framework for the simulation of CR-

VANET channel sensing, coordination and allocation," Ad-Hoc and Sensor Wireless

Networks, vol. Submitted, 2014.

[118] S. H. Abbassi, I. M. Qureshi, H. Abbasi and B. R. Alyaie, "History Based Spectrum

Sensing in CR-VANETs," EURASIP Journal on Wireless Communications and

Networking, vol. 2015:163, 2015.

[119] T. Issariyakul and E. Hossain, Introduction to Network Simulator NS2, New York:

Springer Science+Business Media, LLC, 2009.

[120] U. T. Malaysia, "Network Simulator 2," 26 8 2003. [Online]. Available:

http://sps.utm.my/wp-content/uploads/2013/12/NS2-Manual2.pdf.

126

[121] School of E. Engineering and C. Science, "CDA6530: Performance Models of Computers

and Networks, Chapter 10: Introduction to NS2," [Online]. Available:

http://www.cs.ucf.edu/~czou/CDA6530-11/NS2-tutorial.pdf.

[122] H. P. HALVORSEN, Introduction to Visual Studio and C#, Telemark: Telemark

University College, 2014.

[123] S. Dhall and P. Sundararaman, "C# and .NET Framework," [Online]. Available:

http://www.cs.colorado.edu/~kena/classes/5448/f12/presentation-materials/dhall.pdf.

[124] D.-G. Wang, W.-Y. Song and L. Zou, "Continuity of Fuzzy Similarity Inference Method,"

Journal of Multiple-Valued Logic and Soft Computing, vol. 18, no. 5-6, pp. 525-539, 2012.

[125] K. D. Singh, P. Rawat and J.-M. Bonnin, "Cognitive radio for vehicular ad hoc Networks

(CR-VANETs): approaches and challenges," EURASIP Journal on Wireles

Communications and Networking, vol. 1, no. 49, 2014.

[126] S. H. Abbassi, I. M. Qureshi and H. Abbasi, "Performance of Uni-directional Road Side

Units in Vehicular Adhoc Networks," in World Congress on Computer Applications and

Information Systems, Hammamet, 2014.

[127] S. H. Abbassi, I. M. Qureshi, H. Abbasi and B. R. Alyaie, "CR Channel Allocation Using

Fuzzy Logic in VANETs," MAGNT RESEARCH REPORT, vol. 3, no. 3, pp. 898-909,

2015.

[128] S. N. Sivanandam, S. Sumathi and S. N. Deepa, Introduction to Fuzzy Logic using

MATLAB, Heidelberg: Springer, 2007.

[129] S. Pandey and P. Agrawal, "A Survey on Localization Techniques for Wireless Networks,"

Journal of the Chinese Institute of Engineers, vol. 29, no. 7, pp. 1125-1148, 2006.

Date post:	26-Oct-2021
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

STRU TUR , OOR INATION, S NSING AN ALLO ATION IN OGNITIV ...

Documents