University of Windsor

Scholarship at UWindsor

Electronic Theses and Dissertations Theses, Dissertations, and Major Papers

11-7-2015

Architecture for Extracting Data from Vehicular Sensors

Gaurav Sood
University of Windsor

Follow this and additional works at: https://scholar.uwindsor.ca/etd

Recommended Citation
Sood, Gaurav, "Architecture for Extracting Data from Vehicular Sensors" (2015). Electronic Theses and
Dissertations. 5514.
https://scholar.uwindsor.ca/etd/5514

This online database contains the full-text of PhD dissertations and Masters’ theses of University of Windsor
students from 1954 forward. These documents are made available for personal study and research purposes only,
in accordance with the Canadian Copyright Act and the Creative Commons license—CC BY-NC-ND (Attribution,
Non-Commercial, No Derivative Works). Under this license, works must always be attributed to the copyright holder
(original author), cannot be used for any commercial purposes, and may not be altered. Any other use would
require the permission of the copyright holder. Students may inquire about withdrawing their dissertation and/or
thesis from this database. For additional inquiries, please contact the repository administrator via email
([email protected]) or by telephone at 519-253-3000ext. 3208.
Architecture for Extracting Data from Vehicular Sensors

By

Gaurav Sood

A Thesis
Submitted to the Faculty of Graduate Studies
through the School of Computer Science
in Partial Fulfillment of the Requirements for
the Degree of Master of Science
at the University of Windsor

Windsor, Ontario, Canada

2015

© 2015 Gaurav Sood

Architecture for Extracting Data from Vehicular Sensors

by

Gaurav Sood

APPROVED BY:

______________________________________________
Dr. Huapeng Wu
Department of Electrical and Computer Engineering

______________________________________________
Dr. Jianguo Lu
Department of Computer Science

______________________________________________
Dr. Ataul Bari, Co-advisor
Department of Electrical and Computer Engineering

______________________________________________
Dr. Arunita Jaekel, Advisor
Department of Computer Science

August 21, 2015

 DECLARATION OF ORIGINALITY

I hereby certify that I am the sole author of this thesis and that no part of this
thesis has been published or submitted for publication.

I certify that, to the best of my knowledge, my thesis does not infringe upon
anyone’s copyright nor violate any proprietary rights and that any ideas, techniques,
quotations, or any other material from the work of other people included in my thesis,
published or otherwise, are fully acknowledged in accordance with the standard
referencing practices. Furthermore, to the extent that I have included copyrighted material
that surpasses the bounds of fair dealing within the meaning of the Canada Copyright Act,
I certify that I have obtained a written permission from the copyright owner(s) to include
such material(s) in my thesis and have included copies of such copyright clearances to my
appendix.

I declare that this is a true copy of my thesis, including any final revisions, as
approved by my thesis committee and the Graduate Studies office, and that this thesis has
not been submitted for a higher degree to any other University or Institution.

iii
 ABSTRACT

In this thesis we investigate an alternate source of vehicular information for

collision avoidance systems and driver assistance applications, which is more

accurate, reliable in all conditions and has minimum time lag. We have designed

and developed an architecture, which enables us to read, analyze, decode and store

the real-time vehicular data from the vehicle’s electric sensors. We have designed

two algorithms for decoding the raw data read from the vehicle’s Controller Area

Network (CAN) [7] bus, to which various electric components of the vehicle are

connected to communicate the real-time data.

We have shown that the vehicular speed which is a very important

parameter in the calculation of ‘Time to Collision (TTC)’ by collision avoidance

algorithms [22] is more accurate, reliable and has higher polling rate, when

calculated from the vehicle’s CAN bus as compare to the other source of

information i.e. GPS [6].

iv
 DEDICATION

To my father whose shower of blessings remains with me and always motivates

me to accomplish the ambitions he had seen for me.

v
 ACKNOWLEDGEMENTS

My ultimate gratitude goes to my supervisor Dr. Arunita Jaekel, for

providing me the opportunity to work in an exciting and challenging field of

research. Her constant support, guidance and belief in me have played a major role

towards successful completion of my thesis. Her suggestion to work as a Research

Intern at Arada Systems, Windsor proved decisive in defining my research goals

and in identifying the ways to achieve them.

I would like to acknowledge Dr. Ataul Bari, Dr. Huapeng Wu and Dr. Jianguo Lu

whose suggestions and recommendations greatly improved the quality of this

work. I would also like to thank Mr. Nazeer Shaik for his suggestions and

arrangements to complete the testing in Michigan, USA. I am thankful to my

friend and colleague Jordan Willis for his continuous support and help in this work

and learning new things. And I want to thank my best friends Arshdeep Singh

Sidhu and Manpreet Kaler for always standing besides me and motivating me in

tough times.

Finally, to three wonderful women in my life, my caring mother, my loving

and supportive wife and my sweet sister: my deepest gratitude. Their

encouragement and support are much appreciated during entire phase of my

research work and life in general.

vi
 TABLE OF CONTENTS

DECLARATION OF ORIGINALITY .............................................................................. iii

ABSTRACT....................................................................................................................... iv

DEDICATION .....................................................................................................................v

ACKNOWLEDGEMENTS ............................................................................................... vi

LIST OF TABLES ...............................................................................................................x

LIST OF FIGURES ........................................................................................................... xi

LIST OF APPENDICES .................................................................................................. xiii

LIST OF ABBREVIATIONS .......................................................................................... xiv

CHAPTER-1 INTRODUCTION .......................................................................................1

1.1 Overview ..................................................................................................................... 1
1.2 Motivation ................................................................................................................... 2
1.3 Problem Statement....................................................................................................... 3
1.4 Solution Outline .......................................................................................................... 4
1.5 Thesis Organization ..................................................................................................... 6

CHAPTER-2 REVIEW OF LITERATURE ......................................................................7

2.1 Introduction to Intelligent Transportation Systems ............................................................. 7
2.2 Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V) Communication................. 9
2.2.1 Dedicated Short Range Communications Protocol .................................................... 11
2.2.2 V2V Devices ............................................................................................................ 13
2.2.2.1 Original Equipment Manufacturer (OEMs) ............................................. 14
2.2.2.2 Aftermarket Devices ............................................................................... 14
2.2.2.3 Infrastructure Based Devices ................................................................... 16
2.2.3 V2V Safety Applications .......................................................................................... 16
2.3 Controller Area Network (CAN) Bus............................................................................... 17
2.3.1 CAN Frame Structure Format ................................................................................... 20

vii
 2.3.2 CAN Bus Data Extraction ......................................................................................... 22
2.4 Collision Avoidance Systems .......................................................................................... 24
2.4.1 Global Positioning System Data Based Systems ........................................................ 24
2.4.2 Other Novel Techniques ........................................................................................... 26

CHAPTER-3 .....................................................................................................................30

THESIS OBJECTIVE AND RESEARCH METHODOLOGY ........................................30

3.1 Introduction..................................................................................................................... 30
3.2 Synopsis of Problems and Limitations ............................................................................. 31
3.3 Statement of Objectives ................................................................................................... 32
3.4 Research Methodology .................................................................................................... 33
3.4.1 Architecture of the System ........................................................................................ 35
3.4.2 Software Architecture ............................................................................................... 39
3.4.3 Software UML Class Structure ................................................................................. 45

CHAPTER-4 RESULTS AND ANALYSIS ....................................................................52

4.1 Test Setup ....................................................................................................................... 52
4.2 Comparison of CAN Speed and GPS Speed..................................................................... 53
4.2.1 Test Results at a Parking Lot..................................................................................... 53
4.3.2 Test Results at a City Road ....................................................................................... 55
4.3.3 Test Results on a Highway........................................................................................ 57
4.3.4 Test Results at Downtown ........................................................................................ 58
4.3 Other Vehicular Sensor’s Data ........................................................................................ 59

CHAPTER-5 CONCLUSION & FUTURE WORK ........................................................61

5.1 Conclusion ...................................................................................................................... 61
5.2 Future Work .................................................................................................................... 62

REFERENCES.................................................................................................................63

APPENDICES ..................................................................................................................69
Appendix A .......................................................................................................................... 69
Big Endian Algorithm ........................................................................................................... 69
Appendix B........................................................................................................................... 71
Little Endian Algorithm ........................................................................................................ 71

viii
VITA AUCTORIS ...........................................................................................................73

ix
 LIST OF TABLES

Table I: Typical DSRC data traffic requirements…………………………….......13

Table II: Aftermarket Safety Device Types……………………………………....15

x
 LIST OF FIGURES

Figure 2.1: National ITS Architectures of America………………………………..8

Figure 2.2: Framework for integrated transportation system………………………8

Figure 2.3: V2I controlled traffic situations………………………………………..9

Figure 2.4: Visual Representation of V2V Communication……………………...11

Figure 2.5: Examples of Crash Scenarios and Vehicle-to-Vehicle Applications…17

Figure 2.6: Structure of CAN Network…………………………………………...19

Figure 2.7: Structure of CAN frame………………………………………………21

Figure 2.8: USB CAN smart card………………………………………………...23

Figure 2.9: CAN frames display window…………………………………………23

Figure 2.10: Graphical illustration of two vehicles A and B and their collision

path………………………………………………………………………………..25

Figure 2.11: Graphical illustration of rear-end collision………………………….26

Figure 2.12: Path prediction using DS reasoning system…………………………28

Figure 3.1: System Architecture Flow Diagram………………………………….34

Figure 3.2: System Architecture………………………………………………….35

Figure 3.3: CAN to OBD-II Cable………………………………………………..36

Figure 3.4: OBU (LocoMate), Arada Systems, Windsor…………………………37

Figure 3.5: GPS Antenna…………………………………………………………38

Figure 3.6: CAN-Ethernet Gateway Flow Diagram………………………………39

Figure 3.7: Screenshot of CAN DBC file………………………………………...40

Figure 3.8: Software Architecture Flow Diagram………………………………...42

Figure 3.9: Screenshot of Raw data ………………………………………………44

xi
Figure 3.10: Screenshot of Sensor’s Decoded Values…………………………....44

Figure 3.11: CAN-Ethernet Application Software UML Class Diagram………...46

Figure 4.1: Comparison CAN Speed and GPS Speed at a Parking Lot…………..54

Figure 4.2: Comparison of CAN Speed and GPS Error at a Parking Lot………...55

Figure 4.3: Comparison CAN Speed and GPS Speed at a City Road…………….56

Figure 4.4: Comparison of CAN Speed and GPS Error at a City Road…………..56

Figure 4.5: Comparison CAN Speed and GPS Speed on a Highway…………….57

Figure 4.6: Comparison of CAN Speed and GPS Error on a Highway…………..58

Figure 4.7: Comparison CAN Speed and GPS Speed at Downtown……………..58

Figure 4.8: Comparison of CAN Speed and GPS Error at Downtown…………...59

xii
 LIST OF APPENDICES

Appendix A: Big Endian Algorithm……………………………………………...69

Appendix B: Little Endian Algorithm……………………………………………71

xiii
 LIST OF ABBREVIATIONS

1. ABS Anti-lock Braking System

2. ACK Acknowledge

3. ARP Address Resolution Protocol

4. ASC Automatic Stability Control

5. AT Auto Tracker

6. BSM Basic Safety Message

7. CA Constant Acceleration

8. CAN Controller Area Network

9. CRC Cyclic Redundant Check

10. CTR Constant Turn Rate

11. CTRA Constant Turn Rate and Acceleration

12. CV Constant Velocity

13. DHCP Dynamic Host Configuration Protocol

14. DLC Data Length Code

15. DS Dempster-Shafer

16. DSRC Dedicated Short Range Communications

17. ECU Engine Control Unit

18. EEBL Emergency Electronic Brake Light

19. EMC Electromagnetic Compatibility

20. EMI Electromagnetic Interference

21. EoF End of Frame

22. ETC Electronic Toll Collection

xiv
23. FCC Federal Communications Commission

24. FIFO First In First Out

25. GPS Global Positioning System

26. ICMP Internet Control Message Protocol

27. IDE Identifier Extension

28. IFS Intermission Frame Space

29. IP Internet Protocol

30. IPT Inverse Perspective Transform

31. ISO International Organization for Standardization

32. ITS Intelligent Transportation System

33. LAN Local Area Network

34. LIDAR Light Detection and Ranging

35. LRR Long Range Radar

36. NHTSA National Highway Traffic Safety Administration

37. OBD On-board Diagnostics

38. OBU On Board Unit

39. OEM Original Equipment Manufacturer

40. RSE Road Side Equipment

41. RTR Remote Transmission Request

42. SoF Start of Frame

43. SpaT Signal Phasing and Timing

44. TCP Transmission Control Protocol

45. TTC Time to Collision

xv
46. UDP User Datagram Protocol

47. UKF Unscented Kalman Filter

48. USB Universal Serial Bus

49. V2I Vehicle to Infrastructure

50. V2V Vehicle to Vehicle

51. VANET Vehicular ad-hoc Network

xvi
 CHAPTER-1

INTRODUCTION

1.1 Overview

Vehicular safety technology has developed rapidly in the last two decades.

Modern vehicles can protect occupants much better in the event of a crash, due to the

advanced structural design techniques complemented by more stringent road standards,

and some mandatory standard crash avoidance technologies. All these advances have

made motor vehicles safer than they have ever been. However, there are still a significant

number of annual crashes that could potentially be eliminated through expanded use of

more advanced crash avoidance technologies. It has been estimated by National Highway

Traffic Safety Administration (NHTSA) [5] that there are more than five million annual

vehicle crashes with property damage, injuries and fatalities. So, if deployment of

technology can help to avoid these crashes, the damage due to crashes never occurs.

Some of the most advanced collision avoidance technologies present on the

vehicles today include cameras, on-board sensors and radar applications. These

technologies may warn the drivers of impending collision thus helping them to take

corrective actions. While these current technologies are highly beneficial, the Vehicle to

Vehicle (V2V) communications [4] represent an additional step in helping to warn

drivers of the impending collision threats. V2V communications use on-board dedicated

short-range radio communication (DSRC) [8] [9] devices to transmit messages about a

vehicle’s location information, speed, heading, brake status, and other information to

other vehicles. Information is received in the form of messages, with capabilities of range

1
and line of sight, which can be more than twice of that of the current systems. This longer

detection distance and ability to see around corners or through other vehicles helps V2V-

equipped vehicles perceive some threats sooner than sensors, cameras, or radar can, and

warn their drivers accordingly. V2V can be combined with other driver assistance

applications to act as a complete system that cover the scenarios which are not covered

by V2V alone, such as lane or road departure warning. A complete system could also

augment system accuracy, potentially leading to improved warning timing and reducing

the number of false warnings [5].

1.2 Motivation

According to the statistics published by NHTSA, a total number of 99,176

fatalities was reported in the United States, due to the vehicle traffic crashes during the

period of 2012-2014, with the year 2012 having the maximum number of fatalities at

33,782 [47]. According to another data record published by NHTSA, more than 10

million motor vehicle accidents were reported every year in the US and economic cost of

traffic crashes up till year 2010, was estimated to USD 242 billion [48] [49]. According

to the ‘2013 National Statistics’ given by NHTSA, total number of 30,057 fatalities and

1,591,000 injuries was reported due to road accidents in the US. These vehicle traffic

crashes had also resulted in a property damage of USD 4,066,000 [49]. According to the

‘Canadian Motor Vehicle Traffic Statistics 2012’, total number of 123,963 vehicular

collisions was reported and the related fatalities and serious injuries reported was 2077

and 10,655 respectively [50].

2
 Researchers and technology experts from around the world are working to

develop new ways to minimize the losses occurring due to traffic collisions. Various

technologies such as Airbag System, and Anti-lock Braking System have been

incorporated and mandated by vehicle manufacturers and traffic regulatory bodies over

the period of time, which have proved highly beneficial to reduce the number of vehicle

crashes and related losses. To add more sophistication, a lot of work is being carried on

the improvement and development of collision avoidance algorithms. These algorithms

incorporated with V2V communications are considered as very useful technologies to be

incorporated in the next generation of vehicles. The accuracy of these algorithms depends

on the availability of real time, reliable and accurate source of input parameters e.g.

positioning information, speed and brake status. In this thesis, we focus on techniques

for accurate and timely collection of various vehicular parameters (e.g. speed, engine

rpm, brake status etc.), which can aid in the development of better collision avoidance

algorithms and driver assistance applications.

1.3 Problem Statement

Currently, most of the collision avoidance algorithms rely on GPS [6] for the real-

time situation parameters of the vehicles. For example, the collision warning system

presented in [22] and [23], calculates ‘Time to Collision (TTC)’ between two vehicles

using GPS information (e.g. longitude, latitude, speed, heading, altitude and acceleration)

exchanged using V2V communications. Another system presented in [32], uses vehicular

ad-hoc network (VANET) router, a GPS system, inertial sensors for measuring the speed,

acceleration, and yaw rate and a long-range radar (LRR) to predict the future path of the

3
vehicle. Since these systems work in real time conditions, so the input parameters need to

be accurate, reliable in all conditions and have minimum time lag.

The data obtained from GPS signals is significantly less accurate during bad

weather conditions, in urban downtown areas and tunnels, either because of the signal

distortion, or no GPS signals are received in these conditions. Also, the polling rate of

current GPS receivers is low with intervals up to 200 milliseconds, which may not be

adequate for real-time systems like collision avoidance systems. Similarly, a system

based on camera would have high maintenance costs and also may give less accurate

warnings during bad weather conditions and at night. Furthermore, current source of

information like GPS provides access to only a limited number of vehicular parameters

like speed and acceleration. This has led to the development of collision avoidance

algorithms with limited scope and accuracy. All these factors have resulted in inaccurate

calculation of TTC [22], which resulted in wrong or no warnings of the impending

collisions. So, there is a need to improve the source of vehicular information which

should be more accurate, reliable in all conditions and has higher polling rate. Also, a

number of additional important vehicular parameters like steering angle, brake status,

yaw rate, turn indicator status etc. should be extracted, which could lead to the

development of more sophisticated collision avoidance algorithms.

1.4 Solution Outline

We have found that there is a scope to get the information e.g. speed, acceleration

and brake status directly from the vehicle instead of present sources e.g. GPS, while

positioning information like longitude, latitude, altitude and heading still must be taken

4
from GPS. Extracting data directly from vehicle’s sensors would result in more accurate

and real-time data with very less time lag. In modern vehicles, all the electric devices of

the vehicle are connected to each other by a bus network called Controller Area Network

(CAN) [7]. The vehicle’s electric devices communicate with each other through CAN

protocol, which helps the devices to access the most recent information from various

components. Each message has a unique CAN ID present in the header which helps the

devices to recognize different messages.

The CAN bus can be accessed using the OBD-II port [19] present in the vehicle.

We have designed and developed a software application named CAN-Ethernet which

uses an Ethernet dongle to connect to the OBD-II port of the vehicle. The CAN Ethernet

software reads, analyzes, decodes and stores the CAN data read from the CAN bus in

real-time. The software runs as a server application to act as a source of CAN data and

enables up to five different client connections at the same time. Can-Ethernet software

takes, as input, a DBC file [44], which defines the CAN IDs and other information

helpful in decoding the CAN messages. We have designed two algorithms, one for little

endian [51] and other for big endian byte order [51], that decode the CAN messages read

from the CAN Bus. After decoding, the information is populated in data structures that

store the real time value of each sensor parameter and also written to the log files. We are

able to extract a number of vehicular parameters like speed, engine rpm, brake status,

steering angle, yaw rate, accelerator pedal position, headlight status, wiper status,

ambient temperature and ambient pressure. All these parameters information can prove

very useful in developing and improving collision avoidance systems and driver

assistance applications.

5
1.5 Thesis Organization

The remainder of our thesis is organized as follows. Chapter 2 gives the brief

description about the Intelligent Transportation Systems (ITS) including V2V and V2I

technologies. A brief introduction about CAN Bus protocol and some existing collision

avoidance technologies is also provided in this chapter. Chapter 3 presents our proposed

CAN-Ethernet application software architecture. Chapter 4 discusses and analyzes our

test results. Chapter 5 concludes our work and identifies some possible directions for the

future work.

6
 CHAPTER-2

REVIEW OF LITERATURE

2.1 Introduction to Intelligent Transportation Systems

“Intelligent transportation systems (ITS) are advanced applications which,

without embodying intelligence as such, aim to provide innovative services relating to

different modes of transport and traffic management and enable various users to be better

informed and make safer, more coordinated, and 'smarter' use of transport networks” [1].

In general, ITS is an application of advanced technologies which includes computers,

sensors, communications, and various electronic devices to make commutation safer,

quicker, energy efficient and environment friendly.

The growing congestion and safety problems in transportation networks along

with infrastructural, environmental and land constraints have resulted in the demand of

adequate and efficient use of the existing resources. The application of ITS is critical in

solving such problems. Various sophisticated applications are already in use and are

being developed to improve transportation around the world. For example, computers and

communication systems are being integrated to provide link between vehicles, travelers

and infrastructures to address the modern problems of traffic collisions and congestion to

save precious lives and energy.

National ITS architectures for several countries were designed and planned,

including the research area and transportation service. The American national ITS

Architectures is shown in Figure 2.1.

7
Figure 2.1: National ITS Architectures of America [2] [3]

A basic framework for developing an integrated transportation systems provided

by ITS architecture is shown in Figure 2.2.

Figure 2.2: Framework for integrated transportation system [3]

8
2.2 Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V)
Communication

In V2I, the infrastructure plays a coordination role by gathering situation

awareness information on traffic and road conditions and then suggesting or imposing

certain behavior conditions on the vehicles. In other scenarios, vital information about

traffic signals or pedestrian movement is transmitted to the vehicles thereby alerting the

drivers to react in a timely fashion. In a more sophisticated scenario, the velocities and

accelerations of vehicles and inter-vehicle distances would be suggested by the

infrastructure on the basis of traffic conditions, with the goal of optimizing overall

emissions, fuel consumption, and traffic velocities. The information can be broadcast to

the vehicles or can be relayed to other vehicles by using multi-hop networking of the

V2V.

Figure 2.3: V2I controlled traffic situations [4]

9
 The figure 2.3 shows two different traffic situations [4]. In the left panel, traffic

density is low and the central infrastructure based controller acts to improve fuel

efficiency and reduce emissions of individual vehicles, smoothing accelerations and

decelerations; in the right panel, due to greater congestion, the infrastructure control is

primarily concerned with depleting queues at intersections with an eye toward global fuel

economy and emissions reduction.

V2V is a decentralized system as compared to V2I, which aims at organizing the

interaction among vehicles and possibly developing collaborations among them [4]. In

this system, vehicles periodically exchange information to make decisions for the future

behavioral actions. In a general V2V scenario, when two or more vehicles come within a

communication range, they connect automatically and make an ad-hoc network to share

their position, direction, speed, acceleration, braking and various other parameters’ data.

The vehicles get their own parameters’ data from various sources like Global Positioning

Data (GPS) [6], Controller Area Network (CAN) [7], and Road Side Equipments (RSEs)

[5]. The devices placed on the vehicles to receive, process and transmit the data are

called On-Board Units (OBUs) [5]. These OBUs act as routers which enable the multi-

hop transmission of the data to more distant vehicles and road side stations.

Figure 2.4 shows a visual representation of a V2V communication network.

10
Figure 2.4: Visual Representation of V2V Communication [5]

The wireless communication is based on IEEE 802.11a, also known as Dedicated

Short Range Communications (DSRC). A frequency spectrum of 5.9-GHz range has been

allocated to support the development of safety applications based on V2I and V2V

communication systems [5] [8].

2.2.1 Dedicated Short Range Communications Protocol

Dedicated Short-Range Communications (DSRC) is 75 MHz of spectrum at 5.9

GHz allocated by the Federal Communications Commission (FCC) to “increase traveler

safety, reduce fuel consumption and pollution, and continue to advance the nation’s

economy [8].” It is developed to dedicatedly support V2V and V2I communications

using a variant of IEEE 802.11a technology. DSRC would support safety-critical

11
communications like collision warnings, as well as other valuable ITS applications such

as Electronic Toll Collection (ETC), digital map update, etc. The versatility of DSRC

greatly enhances the likelihood of its deployment by various industries and adaptation by

the consumers.

The 2004 FCC ruling [9] specifies DSRC will have six service channels and one

control channel. The control channel is to be regularly monitored by all vehicles. Safety

messages, whether originated by vehicles or roadside transmitters, are to have priority

over non-public safety communications. Therefore all safety messages are to be sent in

the control channel. In the meantime, a licensed roadside unit could use the control

channel to inform approaching vehicles of its services (often non-safety applications) and

conduct the actual application in one of the service channels. For example, a roadside unit

could announce a local digital map update in the control channel and transfer this data to

interested vehicles in a service channel.

Table I illustrates the sort of DSRC data traffic characteristics being used in the

standards deliberations [10] [11] [12]. The FCC has recognized safety messages and

“safety of life” messages. Safety of life is to have the highest priority. The non-safety

data transfers have the lowest priority. The non-safety communications happen for file

transfers (e.g., infotainment) or transactions (e.g. toll collection). Transactions may

require the exchange of two or three messages within a short time window (e.g. 20 ms).

12
 Table I
Typical DSRC data traffic requirements

Applications Packet Allowable Network Message Priority

Size(bytes)/ Latency Traffic Range
Bandwidth (ms) Type (m)

Intersection Collision ~100 ~100 Event 300 Safety of

Warning/Avoidance Life

Cooperative Collision ~100/ ~100 Periodic 50 - 300 Safety of

Warning ~10 Kbps Life

Work Zone Warning ~100/ ~1000 Periodic 300 Safety

~1 Kbps

Transit Vehicle ~100 ~1000 Event 300 – Safety

Signal Priority 1000

Toll Collection ~100 ~50 Event 15 Non-Safety

Service ~100/ ~500 Periodic 0 - 90 Non-Safety

Announcements 2 Kbps

Movie Download (2 > 20 N/A N/A 0 - 90 Non-Safety

hours of MPEG 1): Mbps
10 min download
time

2.2.2 V2V Devices

Various types of V2V devices are available which are either placed inside the

vehicle during manufacturing or can be installed as an aftermarket product in the vehicle.

13
All these types support different types of V2V and V2I safety applications. These devices

can be categorized as follows.

2.2.2.1 Original Equipment Manufacturer (OEMs)

An OEM device is an electronic device integrated into a vehicle during vehicle

manufacturing. A V2V system is integrated into these devices which is connected to the

proprietary CAN busses [7] and can provide highly accurate information running in these

busses. This vehicle information along with the data collected from the GPS receivers is

used to generate Basic Safety Messages (BSMs) [5]. The integrated system both

broadcasts and receives BSMs and also has a processor that can process the contents of

the received messages to provide advisories and warnings to the driver of the vehicle in

which it is installed. Because the device is fully integrated into the vehicle at the time of

manufacture, vehicles with Integrated Safety Systems could potentially provide haptic

warnings to alert the driver (such as tightening the seat belt or vibrating the driver’s seat)

in addition to audio and visual warnings provided by the aftermarket safety devices. The

equipments required for an integrated OEM V2V system would consist of a general

purpose processor, associated memory, a radio transmitter and transceiver, antennas,

interfaces to the vehicle’s CAN bus, and a GPS receiver. These types of integrated

devices can be collaborated with other vehicle-resident collision avoidance systems to

exploit the functionality of both types of systems.

2.2.2.2 Aftermarket Devices

Automotive aftermarket devices can be defined as any device with one or more

functions in the areas of safety, comfort, performance or convenience, which are added to

14
the vehicle after its original assembly. A V2V communication aftermarket device

generally provides warnings and/or advisories to the driver similar to those provided by

OEM installed V2V devices. These devices, however, may not be as fully integrated into

the vehicle as an OEM device, and the level of connection to the vehicle can vary based

on the type of aftermarket device itself. For example, an independent V2V aftermarket

safety device could only connect to a power source, and otherwise would operate

independently from the systems in the vehicle. Aftermarket V2V devices can be added to

a vehicle at a vehicle dealership, as well as by authorized dealers or installers of

automotive equipment. Some aftermarket V2V devices (e.g., cell phones with apps) are

portable and can be standalone units carried by the operator or the passenger.

Table II [5] shows the details of all types of aftermarket devices that are employed

in the vehicles.

Table II
Aftermarket Safety Device Types

Device type Definition Method of Functionality

Installation
Device is able to be Device would need
connected to the to be installed by a Transmits BSM
vehicle for power certified installer on
Vehicle Awareness source. Device vehicles not
Device provides Basic equipped with V2V
Safety Message for technology to
surrounding ensure correct
vehicles. antenna placement
and security.

Device is connected This device only

to the vehicle for receives power from V2V Safety
power source, the vehicle; applications
Aftermarket Safety Device transmits however, a certified Receives and
Devices BSM and receives installer would need Transmits BSM
(i.e., Self-contained) BSMs to support to ensure correct Driver-Vehicle

15
 safety applications antenna placement Interface
for the driver of the and security.
vehicle in which it
is installed.

Device is connected This device needs to

to the vehicle’s data be connected to the V2V Safety
bus that provides vehicle’s data bus, applications
BSM and safety therefore would Receives and
applications for the require an installer Transmits BSM
Retrofit Safety driver of the vehicle that can access this Driver Vehicle
Devices in which it is for the particular Interface
installed. make of vehicle. Integration into the
Also, a certified vehicle data bus
installer would need
to ensure correct
antenna placement
and security.

2.2.2.3 Infrastructure Based Devices

Infrastructure based devices can be defined as devices that are co-located along

with road side equipments that allow vehicles to receive information from the

infrastructure. The road side equipments can be road signs, traffic signals etc. General

applications that run on these devices are related to signal phasing and timing (SPaT),

curve and curve speed warnings, traffic advisories etc.

2.2.3 V2V Safety Applications

V2V technology communicates via DSRC (radio signals) which offer 360 degree

coverage. This communication allows vehicles equipped with V2V devices to track each

other at times when they can be completely unaware of each other’s presence. DSRC

protocol has an operational range of up to 300 meters that facilitate identification of

intersecting paths that potentially result in a crash if no vehicle takes necessary action. In

16
addition, V2V systems are not affected by weather conditions, light, or conditions of road

as in the case of other systems dependent on cameras or Light Detection and Ranging

(LIDAR) [13].

In figure 2.5 various safety applications and scenarios they can address are shown.

Figure 2.5: Examples of Crash Scenarios and Vehicle-to-Vehicle Applications [5]

2.3 Controller Area Network (CAN) Bus

There are many electronic devices in today’s in-vehicle e.g. anti-lock braking

system (ABS) and automatic stability control (ASC), and the number of these devices

incorporated is increasing gradually. These devices need to communicate with each other.

As a point-to-point communication would increase the wiring that would lead to increase

in the vehicular weight, so the bus network is used for the multi point communication.

17
Controller Area Network or CAN bus designed and developed by Robert Bosch GmBh

[7] is international standardized and defined in the ISO 11898-1 [16]. As a network

specification for electric control system of vehicle, CAN has the capacity of real time

control. The efficient use of these electronic devices through bus network has helped to

reduce the oil consumption and air pollution, improve the ability, safety and comfort of

the vehicle and make drive easy. Some of the electronic devices that are used are engine

timer, auto tracker (AT) and accelerator paddle controller, anti-lock braking system

(ABS), automatic stability control (ASC), intelligent air bag, air conditioning and sound

device [14].

The Controller Area Network (CAN) is a serial communications protocol which

efficiently supports distributed real-time control with a very high level of security [7].

CAN is one of the most popular in-vehicle networks. It simplifies the cable in vehicle, so

the weight of the vehicle is reduced and easy to maintain and repair [15].

In figure 2.6 a general structure of the CAN network is shown. Many electronic

devices are connected to CAN network and communicate with each other. The

microcontroller in each electronic device communicate to CAN controller and CAN

controller connects CAN bus via CAN driver. CAN network is composed of 2 lines and

communicates by using difference of voltage level [7] [15].

18
 Figure 2.6: Structure of CAN Network [15]

The various advantages of CAN protocol are discussed as follows [14] [17] [18].

• Latency Time: This is defined as the time a transmitting node is ready to send

information up until the time the transmission has been completed. CAN has the

short latency time that is essential in the real-time control applications.

• Electromagnetic Compatibility (EMC) and Electromagnetic Interference (EMI):

The performance is not affected by radiated emissions and susceptibility.

• Error Handling and Fault recovery: Less than one undetected error rate in the

lifetime of the vehicle is considered reasonable. The ability to handle and recover

quickly from faults was considered as an important factor for the CAN protocol.

• Data Consistency: Ensure the consistency of data across the network, particularly

when sharing sensor information after eliminating duplications.

• Flexibility: Since vehicle configuration requirements may be different according

to model variants and model years, it is necessary for ECU's to be interconnected

19
 at different locations in the vehicle free from the need to redesign or re-qualify the

system or sub-system.

• Expandability: It is possible for existing systems to be upgraded or added to over

time without modification of the original system if the additional ECU's are

listening nodes.

• Broadcasting: Support a multi-master broadcast communications system. In other

words, every connected control unit has the same right to access the network.

• The protocol can take up to 2032 identifiers and support up to 1Mbps over 40

meters of twisted pair cable.

• Contention based with no data loss. Simultaneous transmissions are resolved via

non-destructive bitwise arbitration.

2.3.1 CAN Frame Structure Format

The CAN protocol supports two message frame formats, the only essential

difference being in the length of the identifier. The “CAN base frame” supports a length

of 11 bits for the identifier, and the “CAN extended frame” supports a length of 29 bits

for the identifier [16].

Figure 2.7 below shows a structure of the CAN data frame [15] [16]. A “CAN

base frame” message begins with the start bit called "Start Of Frame (SOF)", this is

followed by the “Identifier" which is used to specify the “CAN ID” of the sensor and the

"Remote Transmission Request (RTR)" bit used to distinguish between the data frame

and the data request frame called remote frame. If the message is used as a remote frame,

the DLC contains the number of requested data bytes. The following field contains the

20
"Identifier
entifier Extension (IDE)" bit tto
o distinguish between the CAN base frame and the CAN

extended frame, as well as the "Data Length Code (DLC)" used to indicate the number of

following data bytes in the "Data field". The "Data field" that follows is able to hold up to

8 data byte. The integrity of the frame is guaranteed by the following "Cyc
"Cyclic Redundant

Check (CRC)" sum. The "Acknowledge

"Acknowledge (ACK) field" compromises the ACK slot and the

ACK delimiter. The bit in the ACK slot is sent as a recessive bit and is overwritten as a

dominant bit by those receivers, which have at this time received the data
d correctly. The

end of the message is indicated by "End ooff Frame (EOF)". The "Intermission Frame

Space (IFS)" is the minimum number of bits separating consecutive messages.

Figure2.7: A structure of CAN fram

frame [16]

In the “CAN extended frame” the length of the identifier used is 29 bit. The 29 bit

identifier is made up of the 11-bit

11 t base identifier and 18 bit extension. The distinction

between CAN base frame format and CAN extended frame format is made by using the

IDE bit, which is transmitted as dominant in case of an 11-bit

bit frame, and transmitted as

recessive in case of a 29-bit

29 frame. The two frame formats coexist on the bus at the same

time and “CAN base frame” always has the higher priority over the “CAN extended

frame” in case of collision fo

for the bus access.

There are four different types of frames transmitted in the CAN protocol [7].

1. A data frame carries data from a transmitter to the receivers.

21
 2. A remote frame is transmitted by a bus unit to request the transmission of the data

frame with the same identifier. The "Remote Transmission Request (RTR)" bit is

used to distinguish between the data frame and the remote frame.

3. An error frame is transmitted by any unit on detecting a bus error.

4. An overload frame is used to provide for an extra delay between the preceding

and the succeeding data or remote frames.

2.3.2 CAN Bus Data Extraction

There has not been any published material found explaining the ways for

extracting and decoding the data from the CAN bus. CAN bus monitoring requires a

CAN USB/Ethernet/WiFi device that could be connected to the vehicle’s OBD-II [19]

port. A CAN application running on a computer is connected to this device and CAN

frames are sent and received by this application through the CAN device.

In [20], the CAN bus monitoring is done using an USB-CAN device (Figure 2.8)

and logging the read raw data in text format. The work is limited to passively monitor the

CAN bus using device USB-CAN device without intending to decode the raw values.

The hardware of USB-CAN smart card consists of power module, peripheral electric

circuit, USB interface module and CAN interface module [21].

22
Figure 2.8: USB-CAN smart card [20]

The number of data bytes in the CAN frame received from the bus varies

according to the CAN ID from where the bytes are received. The CAN device also has

the capabilities of filtering and masking the CAN IDs. This helps to get the CAN frames

from only selected CAN IDs which in turn helps to manage the huge of volume of data

coming out. In figure 2.9 below, a sample window of the collected CAN frames is shown.

Figure 2.9: CAN frames display window [20]

23
2.4 Collision Avoidance Systems

A collision avoidance system is a vehicular safety system designed to avoid

crashes or reduce their severity. The main idea of the design of these systems is to detect

other vehicles in the collision range and then send warnings (audio or visuals) to increase

driver’s reaction time or act automatically (like braking or steering) to avoid an imminent

crash. Various kind of detection mechanisms like all weather radar, camera, V2V and

V2I communications are currently used so as to take the precautionary actions before an

accident. Few of the most common new technologies are forward collision warning,

autobrake, lane departure warning, lane departure prevention, adaptive headlights and

blind spot warning.

2.4.1 Global Positioning System Data Based Systems

These types of systems use V2V and V2I communications to periodically

exchange their GPS [6] location and movement information. Every vehicle has its own

GPS receiver through which they get this information at a predefined polling rate.

Vehicles exchange this information in the form of basic safety messages (BSMs) [5].

These messages are periodically transmitted and received by every vehicle using the

DSRC protocol [8] [9]. The various parameters that are included in these messages are

longitude, latitude, heading, altitude, speed, brake status and acceleration. Every vehicle

processes the information received from these messages along with their own current

information to decide whether there is an impending collision.

Fig. 2.10 illustrates how two vehicles A and B deploy the protocol of periodic

information exchange to avoid collision [22] [23]. The two inner circles are collision

24
zones of vehicles of A and B and the outer circle is communication cluster of A.

Periodically vehicle A receives B’s coordinates and it checks if their collision zones

intersect. To do so, A makes use of the distance between the two vehicles (computed by

the coordinates sent by B) and the collision zone radius (sent by B through the

information packet). If the two vehicle’s collision zones intersect, then vehicle A

computes the next information message from vehicle B, so as to formulate vehicle B’s

motion profile and compare it with its own motion profile. After computing that two

vehicles are on collision course and will intersect at a point then vehicle A calculates the

time to collision (TTC) and communicates the same to vehicle B [22] [23].

Figure 2.10: Graphical illustration of two vehicles A and B and their collision path
[22]

25
 In [24], [25] and [26] the positive impacts of V2V communication to avoid the

rear end collisions are discussed. In figure 2.11, a graphical illustration of rear end chain

collision is depicted. If vehicle A makes a sudden stop then seeing the brake lights of

vehicle A, vehicle N3 will apply emergency brakes and may be able to stop in time. But

because of the line of sight limitation of brake light, it is likely that driver of vehicle N6

will not know of the sudden braking of vehicle A unless vehicle N6 brakes. This leads to

the shortening of the reaction time for the driver of N6 and which can lead to a rear end

collision of N6 with N3. But if the braking status of vehicle A is communicated through

V2V communications then all the vehicles at its rear would be able some extra time to

react appropriately. Also, if the vehicles are equipped with automatic braking systems,

then vehicles at rear would be forced to stop before any impending collision. Similarly,

the vehicles in the adjoining lanes would not be allowed to change the lanes at that

moment through the lane change warnings.

Figure 2.11: Graphical illustration of rear-end collision [24]

2.4.2 Other Novel Techniques

As most of the collision avoidance systems rely heavily on the GPS [6] data for

the vehicle’s position and movement detection, these systems suffer from the limitations

of GPS. This has been well known that GPS signals provide significant less accurate

26
information in bad weather conditions, subways, tunnels and city downtowns, so other

novel techniques are being used for vehicle detection in the systems design.

In [27] a new method for detecting front vehicles in urban traffic was proposed. In

this method a camera is installed on the front windshield of the vehicle. An IPT matrix

which describes the relation between the image coordinate and the real world coordinate

under the assumption that vehicles are on flat road was computed. After that vehicle

candidates are extracted using AdaBoost [28]. The selected candidates were verified by

existence of vertical and horizontal edges for more accurate detection results. The

detected vehicle regions were corrected by shadow [29] and edges [30]. In [31] this

proposed vehicle detection technique was further enhanced by adding features of rear-

lights of vehicles and road characteristics. The accuracy of this technique was proved to

be more than 90% when all the features were used together and even proved to be

efficient during the rainy days.

In [32] a novel path prediction technique for the remote vehicle was proposed. In

this method, every vehicle was equipped with a vehicular ad-hoc network (VANET)

router, a GPS system and inertial sensors for measuring the speed, acceleration, and yaw

rate. There was an unscented Kalman filter (UKF) [33] [34] [35] present for filtering all

these values before they are broadcast to the network. Also, there was a long-range radar

(LRR) mounted on the front bumper of the vehicle. Every vehicle transmits its filtered

GPS and sensor information to the other vehicle and the receiving vehicle performs the

data fusion of the information received from the LRR sensor and the VANET. Then path

prediction algorithm based on dynamic motion models for predicting the future path of

vehicles was called. Four motion models - Constant Velocity (CV), Constant

27
Acceleration (CA), Constant Turn Rate (CTR) and, CTRA (CTR + CA) were available.

A Dempster-Shafer (DS) reasoning system [36], [37] was developed, with the task of

finding the most suitable motion model at every processing cycle. The information

sources of this system were the values of yaw rate, acceleration, and the road curvature

extracted from the digital maps. According to these inputs, the output of the system will

be the most probable motion model at the current processing cycle and its corresponding

belief and plausibility values. Figure 2.12 shows the different motion models selected at

different vehicular movements using DS reasoning system.

Figure 2.12: Path prediction using DS reasoning system [32]

In this chapter we have discussed the role of ITS for safer and efficient use of

transportation networks. The incorporation of various V2V devices into the vehicles can

result in avoiding various road accidents thus saving precious lives and money. We have

also discussed the CAN protocol and structure of CAN bus network. The passive

monitoring CAN bus provides us the access to large number of in-vehicle electronic

28
device’s real time data which could be used in various V2V and V2I applications.

Finally, we have discussed the various methods and techniques used for avoiding

different collision scenarios and also predicting the future vehicular paths.

29
 CHAPTER-3

THESIS OBJECTIVE AND RESEARCH METHODOLOGY

3.1 Introduction

Collision avoidance systems and driver assistance applications have been

implemented using different methodologies e.g. V2V communications, sensors data

processing, and image processing and depend on input parameters from various different

sources e.g. GPS, on-board sensors, long range radars, and cameras. The idea behind

using different techniques is to get a system which would have accuracy and reliability in

all conditions. More accurate and reliable systems would not only help to avoid many

road accidents, thus saving lives and money, but also help to keep the environment clean

by burning less fuel. In our research work, we have made an attempt to get a real time

and reliable source of information which would not suffer from the limitations of other

sources of information discussed in section 2.4.

In this chapter, we present the details of our framework. We present an

architecture of a system that enables us to collect various sensor’s data present in the

vehicle itself. We have developed a software that could reside in the OBUs [5] and

communicates with the vehicle’s CAN [7] bus. The application was developed on Ubuntu

[38] platform using C and C++ as languages. Through this architecture we are able to

collect all sensor’s data available in the CAN bus which includes speed, rpm, accelerator

pedal position, brake status, headlights status, wipers status, left right indicator status,

parking lights status etc. Since these parameters are collected in real time from the

vehicle’s CAN bus, they are proved to be more accurate which would help to increase the

30
efficiency of the collision avoidance systems. In addition, the collection of additional

parameters would help to develop or enhance the driver assistance applications like lane

changing warning system. Also, these parameters are not affected by any changes in the

conditions external to the vehicle.

3.2 Synopsis of Problems and Limitations

We have studied various sources of information used for collision avoidance

systems and driver assistance systems like GPS, LRR or front and rear cameras. Since

these systems work under real time conditions, they have the requirement of real time

data with minimum lag. We have found that all of these sources have their limitations

that reduce their usefulness for developing systems reliable in all weather and road

conditions. For example, a system based on only GPS data would not be reliable in bad

weather conditions, urban downtown areas and tunnels because of the distorted or no

GPS signals received in these conditions. In addition the GPS signals have low polling

rate of up to 200 milliseconds, which should be improved for real time applications like

collision avoidance systems. Similarly, a system based on cameras would suffer from

inaccuracies during bad weather conditions and at night. In addition it would incur high

maintenance costs. An inaccurate system would either give wrong warnings to the driver

or would not give a warning for an impending collision. We have also found that all these

sources of information provide limited access to the full set of parameters that can be

used for these systems. Various vital vehicular parameters like steering angle, accelerator

paddle position, brake status, headlights status, left right indicators etc which could

further enhance the performance of these systems are not available from these sources of

information.

31
3.3 Statement of Objectives

The primary objective of our research is to come up with an architecture of a

system and a generic application, which would enable us to communicate with the

vehicular CAN bus so as to read, decode and analyze the real time data from various

vehicular electronic devices/sensors. This would help us to have a reliable and more

accurate source of information for various collision avoidance systems and driver

assistance systems. We have the goal to provide an alternative to existing sources of

information like GPS because of their limitations in different conditions. For example,

the parameter vehicular speed is required as an input by all of the collision avoidance

algorithms, and is currently taken from GPS data. This parameter’s data can be taken

directly from the vehicle’s CAN bus which would help to remove the lags from speed

taken from the GPS signals. Use of CAN bus data also ensures the continuous feeding of

the updated information due to the high polling rate of 5-7 ms.

In this thesis research, we propose a generic system, which can be adapted for any

vehicle and also would able to act as a common source of vehicular information for any

number of driver assistance applications. The proposed system should have the following

capabilities:

a) Ability to read and decode various vehicular parameters in real time and to test

their accuracy.

b) Ability to test the accuracy of vehicular speed against the speed given by the GPS

signals at the same time. An accurate and more frequent source of information

32
 would certainly help to improve the calculation of the TTC [22] by the collision

avoidance algorithms.

c) Ability to test the accuracy of brake status and the steering angle, which would be

used in the rear end collision warning systems and lane changing warning systems

respectively.

d) Ability to read and decode additional parameters like engine rpm, accelerator

pedal position, yaw rate, headlights, wipers status, engine ambient temperature

and engine pressure which are currently not being used as inputs in the collision

avoidance algorithms and driver assistance applications due to their non-

availability. The availability of these new parameters would help to improve the

accuracy of the current systems and also would lead to the development of

various other applications.

3.4 Research Methodology

The flow diagram of the architecture of our research methodology is depicted in

figure 3.1. In our research, we have developed an application software named ‘CAN-

Ethernet’ which resides in an OBU and makes a TCP [39] [40] connection with an

Ethernet dongle that is connected to the CAN bus of the vehicle. The application software

is designed to send specific commands to the Ethernet dongle, which are interpreted to

the CAN specific format and used to requests raw data from the CAN bus. After

receiving a legitimate command, electronic devices connected to the CAN bus respond

by sending the requested data. The Ethernet dongle passes the received data to the CAN-

Ethernet application software which is waiting for the response. After receiving the raw

data, application analyzes the received raw data at bit level, and decodes it to a human

33
readable form. The decoded data is then made available for collision avoidance

algorithms and other applications residing in the same OBU. Also, it is made available to

be transmitted to the other vehicles in the form of BSMs using DSRC antennas.

Figure 3.1: System Architecture Flow Diagram

The data read from the CAN bus is huge in volume, with a maximum data rate up

to 320 KB/sec, and a maximum latency of < 120 µs for high priority messages. Since a

huge volume of data is available in the CAN bus, the CAN-Ethernet application software

was designed to read, analyze, decode and store the data in parallel so as not to get

crashed while running.

34
3.4.1 Architecture of the System

The architecture of our system is generic and is applicable for any modern vehicle

which has its electronic devices connected to the CAN bus. A vehicle is fitted with the

GPS antenna to receive the GPS signals for longitude, latitude, heading etc information.

This GPS antenna is connected with an OBU which is responsible for the all the

calculations from the raw data, transmission and receiving of DSRC packets. The OBU is

also connected to the Ethernet dongle with a CAT 5e cable [41]. Our CAN-Ethernet

application software makes a TCP connection with the Ethernet dongle using this cable.

The Ethernet dongle is connected to the CAN bus using CAN to OBD-II cable and helps

to communicate with the CAN bus to collect the vital real time vehicular parameter’s data

like speed, steering angle, brake status etc.

Figure 3.2: System Architecture

In figure 3.2 the architecture of the system is shown, which incorporates:

• CAN Bus: The vehicular bus to which all the electronic devices of a vehicle are

connected to communicate with each other as discussed in section 2.3.

35
• Ethernet dongle: We have used an Ethernet dongle similar to the one discussed

in section 2.3.2. The dongle has dimensions of 74 mm x 80 mm and operates on a

5V DC power supply. It has one Ethernet port to connect with a LAN cable and

one RS232 port to make a serial connection. It has one CAN port to connect with

the vehicle’s CAN bus using CAN to OBD-II cable.

• CAN to OBD-II Cable: The cable as shown figure 3.4 is used to provide CAN

signals from the vehicle to the CAN dongle. It has a 16-pin OBD-II male

connector and a 9-pin D-Sub socket female in accordance with CiA 102 [16].

Figure 3.3: CAN to OBD-II Cable

• OBU: We have used an OBU shown in figure 3.5 named LocoMate, marketed by

Arada Systems, Windsor. This device is integrated with GPS, Bluetooth, and high

power 802.11p radios.

36
 Figure 3.4: OBU (LocoMate), Arada Systems, Windsor [42]

• CAT 5e Cable: Cat 5e cable has four twisted pairs in a single cable jacket. It’s an

unshielded twisted cable pair designed specifically for high signal integrity. It is

commonly used for 100Mbit/s networks, but with IEEE 802.3ab defined standards

can also support 1000BASE-T – gigabit Ethernet [41].

• GPS Antenna: The GPS antenna we have used shown in figure 3.7 draws about

10 mA and gives an additional 28 dB of gain. It is magnetic so it sticks to the top

of vehicle.

37
 Figure 3.5: GPS Antenna [43]

The communication with CAN bus through Ethernet port of our dongle is shown

in figure 3.8. The dongle we used, also supports RS232 line to make a serial connection

with the CAN bus. However, the communication path we followed in our CAN-Ethernet

application software is through Ethernet port which is highlighted in Gray and dotted

boxes. The TCP connection is initialized using the IP address of the dongle, which is

provided through our CAN-Ethernet application software. After the connection is

successfully established, a formatted command packet for requesting the CAN data is

sent to the CAN-Ethernet, which is interpreted by the Command Interpreter. The

Command Interpreter validates the received packet, extracts the required data bytes and

sends it to Protocol Converter. The Protocol Converter converts the received bytes to the

CAN protocol format and sends it to the CAN bus through CAN S/W and CAN H/W data

38
buffers. The CAN bus responds to the sent command with the sensor’s raw data, which

are read by the Protocol Converter. The Protocol Converter converts the received bytes to

the dongle’s packet format and sends it to the Command Interpreter through the RS232

data buffer. The Command Interpreter adds the identifying bytes to the packet and sends

it to the CAN Ethernet. Our CAN-Ethernet application software which was continuously

waiting for the response packets, reads, validates and then decodes the received packets.

Figure 3.6: CAN-Ethernet Gateway Flow Diagram

3.4.2 Software Architecture

CAN-Ethernet application software is developed in C /C++ language on Ubuntu

platform. The application takes three inputs as arguments: IP address of the dongle,

39
vehicle’s CAN DBC file [44] and Masks and Filters file. The DBC file, which is a

proprietary format of Vector Informatik GmbH [44], is unique for every vehicle model. A

DBC file contains all electronic devices data decoding information including the CAN ID

assigned to those devices at the time of manufacturing, for a particular vehicle. In figure

3.9, a screenshot of CAN DBC file is shown, which gives the information about the CAN

ID and expected length of data frame for a vehicular sensor in the line starting with

‘BO_’. The following lines starting with ‘SG_’ gives the information about all the signals

expected in the received data frame. The information includes Start bit, Length, Byte

order (Endianness), Value type (Signed or Unsigned), Scale, Offset, Range (minimum

and maximum value) and Units.

Figure 3.7: Screenshot of CAN DBC file [44]

In CAN-Ethernet application software, we parse DBC file information and store it

in a data structure to be able to use it for decoding purposes. We also parse the third

argument of our application which is Masks and Filters file. The masking and filtering

information gives us the capability to request CAN data from only specific vehicle

40
sensors. For example, using masking and filtering we could request the data from only

‘Vehicle Speed’ sensor by setting it to CAN ID = 533. The default mask and filter is set

to value of ‘0’ to get data from all the vehicle sensors. In figure 3.10 the complete flow

diagram of our CAN-Ethernet application software is shown.

41
42
 A

Check
B
Endianness Big Endian

Little Endian

Get start bit Get start bit

and length and length

Little Endian Big Endian

decoder decoder

Final Value = Final Value =

Decoded Decoded
Value * Scale Value * Scale
+ Offset + Offset

Store Final Store Final

Value in Value in
data data
structure structure

End of loop End of loop

Figure 3.8: Software Architecture Flow Diagram

The final decoded value of the vehicle’s sensors is stored in a data structure which

is continuously updated with the new values. This application runs as a server in the OBU

and accepts client connections from the other applications requiring CAN information.

Upon receiving a request our application dumps whole data structure to the client. This

way, it is able to provide real time CAN information to collision avoidance applications

43
and other driver assistance applications running in the same OBU. In figure 3.11 and

figure 3.12, a screenshot of raw data capture and a sensor’s decoded values is shown

respectively.

Figure 3.9: Screenshot of Raw data

Figure 3.10: Screenshot of Sensor’s Decoded Values

44
3.4.3 Software UML Class Structure

The CAN-Ethernet application software is designed using object oriented

approach and the feature of running the application software as a server is added using a

separate thread, thus making it a multi-threaded system. There are eight core classes

designed in the software architecture with additional interface specific classes as

required. The UML class diagram in figure 3.13 shows the object architecture of the

CAN-Ethernet application software.

1) can_main class: The can_main is the main class of the software. It is responsible

for the following activities:

• Validating the three input parameters i.e Ethernet dongle’s IP address,

vehicle specific DBC file and Filters file.

• Initializing the log file based on the current date and timestamp.

• Parsing the Filters file and storing the values in a data structure.

• Initializing the thread of can_server class.

• Initializing the dbc_parser class to parse the input dbc file.

• Initializing the can_utils class.

First the software initializes the can_main class and validates the three

required input parameters. After validating, it checks for the existence of the log

file with the current date and timestamp and creates a new file otherwise. It opens

the Filters file provided as the argument and parses its contents to store in a data

structure present in support_functions. It creates a new thread for can_server class

which runs continuously in a separate thread. It opens an instant of dbc_parser

45
 class passes the DBC file provided as argument to it for parsing. It passes the IP

address of the dongle which is also provided as argument to the can_utils class

instance.

IP Address, vehicle's dbc file

can_main and Filters file must be
-ip : unsigned char present as input
-dbcfilename : char
-Filters : char
can_server * 1 -*cTime::time_t dbc_parser
-bcan_sf : support_functions
+struct CANData() -*dbc : dbc_parser 1 * +*dbc_read_file() : dbc_parser
+main() : int +dbc_write() : void
+Options() : void +struct dbc_t()
1 * +usage() : void *

+show_signals() : void
1

serial_io_tcpip can_utils 1

-host : unsigned char -ip_address : unsigned char

-port : int -port : int
+reconnect() : void -can_sf : support_functions * 1 sensor
+serial_io_tcpip() -can_msg : char -error_count : int
+~serial_io_tcpip() 1 * -*sio : serial_io_tcpip *
+sensor()
+flush() : void -*io : IO_Device
+virtual ~sensor()
+read() : int -*o : OBD2
+virtual get_pid() : unsigned char
+write() : int +write_to_can() : int +virtual get_len() : unsigned char
* +get_value() : bool
* 1

IO_Device
1
* -full_speed : bool
1
-last_msg : double
+IO_Device()
IO_ELM327
+virtual ~IO_Device()
-serial_io_tcptp : serial_io_tcpip +slowdown() : void
-dev_info : unsigned char +virtual flush() : void
+reset_device() : void * * * +virtual send_recv() : void
+recv() : void 1 1
+IO_ELM327() *1
support_functions
+~IO_ELM327()
+flush() : void 1 OBD2
+getWLock() : void +struct filter_list() -*io : IO_Device
+string_reverse() : unsigned char +support_functions()
+determine_supported_pid() : void
+decode() : float +~support_functions() * 1
1 * +OBD2()
+can_decoder() : void +BinaryToInt() : int
+get_filters() : bool
+send_recv() : void +IntToBinary() : unsigned char
+get_pid()
+WriteToSysLog() : void

Figure 3.11: CAN-Ethernet Application Software UML Class Diagram

46
2) can_server class: The can_server class is designed to include the capability of

behaving as server while running. The CAN-Ethernet application can currently

handle up to five client connections simultaneously with capability of more with

some modifications. The main roles and activities of this class are as follows:

• Maintaining a data structure named CANData, with the current

updated decoded values of all the vehicular sensors provided in the

dbc file.

• Creating a connection queue and initializing the read file descriptors to

handle the input from the clients.

• Waiting for the client request and responds by dumping the pointer of

the whole data structure CANData to the client.

3) dbc_parser class: The main purpose of designing dbc_parser class is to get the

useful information present in the dbc file, which is required for decoding the raw

values read from the vehicle’s CAN bus. Following are the main activities of this

class:

• Reading the input dbc file and validates its authenticity.

• Maintaining a data structures for storing the CAN id, sensor name and

expected length of the data field from the CAN bus.

• Maintaining a data structure for storing the signal information in a CAN

id i.e. signal name, start bit, length, byte order (Endianness), value type

(Signed or Unsigned), scale, offset, range (minimum and maximum value)

and units.

• Printing the parsed dbc file information.

47
4) can_utils class: The can_utils class is designed to handle the initialization of

creating the connection to the Ethernet dongle and initializing the instances of the

IO_Device class and OBD2 class. The roles performed by this class are as

follows:

• Responsible for sending the IP address and port number to the

serial_io_tcpip class for creating a TCP connection with the Ethernet

dongle interface.

• Creating an instance of the IO_Device class and then creating new session

of the IO_ELM327 using this instance since IO_Device class was

inherited by the IO_ELM327 class.

• Initializing an instance of OBD2 class and calling sensors class operations

using this instance.

5) serial_io_tcpip class: The serial_io_tcpip class is designed to handle all the

communication with the CAN-Ethernet dongle. The main functions of this class

are as follows:

• Responsible for creating a connection with the Ethernet dongle and

maintaining the created file descriptor and also, trying for reconnection if

connection is not established successfully until time out.

• Writing the command packet to the TCP session using the created file

descriptor which is required to be sent to the CAN bus.

• Reading the raw data from the TCP session using the created file

descriptor from the CAN bus.

48
6) OBD2 class: The OBD2 class is designed to handle the formatting of the

command packet required to be sent to the CAN bus. The Ethernet dongle, we

used, has an interface in it, which accepts command packet in a specific format. It

includes a static PID, length, mask0, mask1, filter0, filter1, filter2, filter3, filter4,

filter5 and checksum. For example, to send a packet to the CAN bus so as to

listen everything over it, we have to format the packet as 0x85, 0x08,

0x00000000, 0x00000000, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x8D. The PID

was set by the manufacturer of this Ethernet dongle so it shouldn’t be changed

while the other fields of the packet can be changed according to the requirements.

7) IO_ELM327 class: This class handles bulk of the work of the CAN-Ethernet

application software. The member functions of this class communicate with the

serial_io_tcpip class member functions to send and receive data byte by byte.

Apart from this, it handles analyzing, decoding and recording of the data read

from the CAN bus. The roles and responsibilities of this class are as follows:

• Reading the formatted packet created in the OBD2 class and writing it to

the CAN bus through serial_io_tcpip class.

• Waiting for the response from the CAN bus and then reading the response

byte by byte.

• Analyzing the each byte read and differentiating it based on the specified

response format.

• Decoding the data bytes based on the information read from the data

structures maintained by the dbc_parser.

49
 • Populating the CANData data structure maintained by the can_server with

the decoded vales of the sensor parameters.

• Recording the decoded values in the created log files.

After writing the command packet on the TCP session, the recv() member

function of the class waits for the CAN bus response. Since, the Ethernet dongle

formatted the response packet before sending it over the TCP session, so the

response packet is expected in a definite format. The first byte is always the

response PID which was set by manufacturer to 0x85. The next byte expected is

length of the data bytes, then the data bytes and in the end checksum byte. After

verifying the checksum, the data bytes are separated from the received packet. In

the data bytes, the first four bytes give the CAN id, which are decoded and

matched with the CAN ids parsed from the dbc file. If a match is found, the byte

order of the signals contained in the data bytes is determined. Two algorithms,

one for decoding based on big endian byte order (Motorola) [51] shown in

Appendix A and other for little endian byte order (Intel) [51] shown in Appendix

B, are designed to decode the data bytes using the start bit and length information

from parsed dbc file. Final value of the raw data is then calculated by applying the

scale and offset information to the decoded value. These final calculated values

are populated in the CANData data structure and logged into the log file. The

CANData data structure always holds the latest values of the sensor parameters as

the old values are continuously replaced with the new values.

8) support_functions class: The support_functions class contains common

functions that are used throughout the entire system. Specific functions that are

50
 commonly used by all objects (for example, BinaryToInt, IntToBinary,

WriteToSysLog etc…) are defined in here. Objects that wish to use the common

function definitions simply just instantiate a support_fucntions object and use its

capabilities.

In this chapter we have discussed the detailed architecture of the system we have

used, describing the physical linking of the various modules of the system. We have

shown and explained the various physical equipments we have used in our research work.

We have also demonstrated the internal communication functionality of the Ethernet

dongle, we have used. We have thoroughly explained the CAN-Ethernet application

software we have developed. We have shown the detailed steps that we had taken to build

this software to be generic and applicable for any vehicle. Finally, we have explained the

algorithms we have designed to decode the raw data from the vehicle’s CAN bus and

populate it in the data structure to be used by collision avoidance algorithms and driver

assistance applications by making client-server connections.

51
 CHAPTER-4

RESULTS AND ANALYSIS

In this chapter, we discuss about our test setup, comparison of the results and

analysis of the data from various vehicular sensors.

4.1 Test Setup

We have done the testing of our software with the cooperation of Arada Systems,

Windsor [42]. We got the access to a testing vehicle, provided by the US Department of

Transportation for testing V2V and V2I related applications, through Arada Systems. The

vehicle we used for our testing is Jeep Grand Cherokee (Year: 2007) and Arada Systems

also helped us to obtain the dbc file of this vehicle model. We have used the Arada

Systems’ OBU (Locomate) as discussed in section 3.4.1, for installing and running our

software application CAN-Ethernet. This OBU is fixed in the provided vehicle and

connected to the Ethernet dongle, which is connected to the OBD-II port of the vehicle’s

CAN bus. The OBU is also connected to a GPS antenna and has the built-in capability of

recording and processing the GPS signals. We have made the adjustments in our software

application, so as to bring the capability of recording the CAN data and GPS data

together into the data log files we are maintaining. We have driven the test vehicle at

various different locations and times to record the CAN and GPS data together. We have

covered parking lots, downtown areas and highways while testing, to check the behaviour

of GPS signals at different locations and different speeds. To do the comparisons between

GPS and CAN data, we parsed the fields collected in the log files into a worksheet

through a custom script program written in Python [45].

52
4.2 Comparison of CAN Speed and GPS Speed

In this section, we have compared the CAN speed against GPS speed at different

locations and the results show that there may be a significant discrepancy between the

actual vehicle speed (based on CAN data) and the speed calculated from GPS signals. In

the remainder of this chapter, we will refer to the speed based on CAN data as CAN speed

and that calculated from GPS signals as GPS speed. We have found that the error in GPS

speed increases during the events of braking or acceleration of the vehicle.

4.2.1 Test Results at a Parking Lot

The first test we conducted was at a parking lot in Troy, USA. This test was done

on a clear sunny day, so we expected comparatively less interference from atmospheric

conditions on GPS signals. In figure 4.1, the CAN speed and the GPS speed are plotted

together on the left y-axis, with total number of records on the x-axis. We have found that

there is a visible deviation of the GPS speed as compared to the more accurate CAN

speed. Also, the deviation is much higher when the vehicle is either accelerating or

braking as compare to when running at a constant speed. Due to the fact that the polling

rate for CAN speed is much higher as compare to the GPS speed, the GPS receivers take

some extra time to recognize the frequent changes in the speed.

53
Figure 4.1: Comparison CAN Speed and GPS Speed at a Parking Lot

We calculated the GPS error in speed by subtracting GPS speed from CAN speed

at a given time and then plotted the calculated GPS error against the CAN speed as

shown in figure 4.2. The CAN speed and GPS error are plotted on the left y-axis and

right y-axis respectively, with total number of records on the x-axis. Left and right y-axis

is used to clearly plot the negative values of the GPS error. We found that the GPS error

has reached up to -7 km/h when the vehicle was braking and coming to stop from a speed

of 20km/h. On other instances, the GPS error ranges from -5 km/h to 5 km/h on certain

changes in speed at different times.

54
Figure 4.2: Comparison of CAN Speed and GPS Error at a Parking Lot

4.3.2 Test Results at a City Road

The second test was conducted at Stephenson Highway, Troy, USA. This test was

conducted on a day with overcast conditions and we have found similar trends in the

deviation of the GPS speed from the CAN speed. The CAN speed and GPS speed are

plotted on the left y-axis and right y-axis respectively, with total number of records on the

x-axis. In figure 4.3, it can be seen that CAN speed curve is more consistent and visibly

moves away from GPS speed curve during acceleration or braking events. The two

curves minimize their gap during constant speed of the vehicle.

55
Figure 4.3: Comparison CAN Speed and GPS Speed at a City Road

In figure 4.4, the calculated GPS error is plotted against CAN speed and we have

found that when the vehicle speed drops from around 60 km/h to 5 kkm/h, the GPS error

reaches to -14 km/h.. Similarly, GPS error reaches close to 8 km/h, when vehicle

accelerates from 10 km/h to 65 km/h. We have also found that, some magnitude of the

GPS error is always present even when the vehicle runs at constant speed.

Figure 4.4: Comparison of CAN Speed and GPS Error at a City Road

56
4.3.3 Test Results on a Highway

The third test was conducted at I-75

75 from Troy to Detroit, USA
USA, on a day with

overcast conditions. In figure 4.5, we plotted the GPS speed and CAN speed, recorded

during the test journey. We have found the two curves almost merge with each other,

when the vehicle was going at around 100 km/h but some deviation is seen
see when the

speed is lowered.

.5: Comparison CAN Speed and GPS Speed on a Highway

Figure 4.5:

In figure 4.6, the calculated GPS error is plotted against the CAN speed. We have

seen that the GPS error is between 0 to -55 km/h when the vehicle was going at around

110 km/h. The GPS error reaches to maximum of -12

12 km/h when the speed is lowered

from around 95km/h to 70 km/h.

km/h

57
Figure 4.6:
.6: Comparison of CAN Speed and GPS Error on a Highway

4.3.4 Test Results at Downtown

The fourth test was conducted in downtown locations at Detroit, USA. In figure

4.7, we have plotted the GPS speed and CAN speed recorded from the test. We have

found that there is a constant deviation of the GPS speed curve from the CAN speed

curve.. Also, while testing we have found that the GPS signals were not consistent due to

the multipath problem [46] of these signals.

Figure 4.7:
.7: Comparison CAN Speed and GPS Speed at Downtown

58
 In figure 4.8, the calculated GPS error is plotted against CAN speed and we have

found a constant error

or of around +5 km/h when the vehicle was accelerating from 0 km/h

to 35 km/h. The GPS error reaches to maximum of -10

10 km/h during the braking event of

the vehicle, when the speed came down from 35 km/h to 0 km/h.

Figure 4.8: Comparison of CAN Speed and GPS Error at Downtown

4.3 Other Vehicular Sensor’s Data

We were also successfully able to get data from various

ious other vehicular sensors

like Engine RPM, Accelerator pedal position, Steering Angle, Brake Status, Headlights

Status, Wipers Status and Yaw rate.

rate. All these parameters can play a very vital role for the

future collision avoidance algorithms and driver assistance applications. As the data from

these sensors is collected in real time ffrom

rom the vehicle’s CAN bus, the accuracy of the

algorithms
ms and applications will increase and, will result in fewer delays.
delays For example,

the current Emergency Electronic Brake Light (EEBL)

( applications rely on either

decrease in the front vehicular speed or image processingg of the front cameras, to

recognize the front vehicle’s braking. But with data available from the CAN bus, the

59
braking information of the front vehicle can be directly transmitted to the rear vehicle,

that will improve the reaction time of the driver in the rear vehicle and will help to avoid

more collisions. Also, the brake status of the front vehicle can be sent directly to multiple

vehicles through multi-hop communication, thus helping to avoid incidents like highway

chain collisions.

In this chapter we have discussed the comparisons of CAN speed and GPS speed

recorded simultaneously during different test drives. We have shown that there is always

some difference between the speed recorded from GPS signals and the speed calculated

from the CAN data. The difference is seen much more prominent and reaches near to 15

km/h on certain occasions of vehicle’s acceleration or braking. We have also found that

the speed calculated from CAN data is more consistent due to the high polling rate of

data from the CAN bus which helps to record even the slight changes in the speed. In

addition, we have also discussed about the number of other vital parameters we

calculated that can help in improving existing collision avoidance algorithms and in

developing new applications.

60
 CHAPTER-5

CONCLUSION & FUTURE WORK

5.1 Conclusion

Due to the increase in the number of the vehicular collisions and loss of lives and

property, a lot of work is being carried out towards the development of the collision

avoidance and driver assistance applications. Researchers from all around the world are

working to develop new techniques and systems, which could improve the existing

algorithms and applications to provide a stable system to the vehicle manufacturers. In

this thesis, we have developed a CAN-Ethernet application software, which will help to

provide a better alternative to existing sources of vehicular information like GPS. We

have shown that the vehicular speed calculated from data read from the CAN bus by our

CAN-Ethernet application is more accurate and consistent as compare to the speed taken

from the GPS signals. This is due to the higher polling rate of data available from the

CAN bus and also due to directly accessing the vehicle’s speed sensor to obtain the

relevant data. We have shown that the errors in the speed given by the GPS signals are

higher during the events of the vehicle’s acceleration and braking. The improvement in

the calculation of the vehicular speed, which is a vital input parameter in the collision

avoidance algorithms, will certainly improve the calculation of the TTC [22]. As we

know that most of the collisions occur either during hard braking of the vehicles or

during sudden accelerations, a more accurate and consistent knowledge of the vehicle’s

speed and, in turn, more accurately calculated TTC will help to generate timely warnings

to the drivers and thereby help to improve their reaction times.

61
 In addition, we are also able to calculate information of various other vital

parameters from the vehicle’s CAN bus like braking status, steering angle, engine RPM,

accelerator pedal positions, yaw rate, headlights status, wipers status, engine ambient

temperature, engine pressure etc. The real time information of all these parameters will

certainly help to improve and develop various collision algorithms and drivers assistance

applications.

5.2 Future Work

In our thesis work, we are able to extract information of various electronic sensors

attached to the CAN bus. The information of some of the parameters was not available

before, so they were not used in the collision algorithms and driver assistance

applications. An important direction for future work would be to develop new collision

avoidance algorithms which utilize these additional parameters’ information. Another

work can be to get access to more CAN IDs of the vehicular sensors and to calculate and

use the additional information.

62
 REFERENCES

1. Official Journal of the European Union, “Directive 2010/40/eu of the European

Parliament and of the Council of 7 July 2010 on the framework for the

deployment of Intelligent Transport Systems in the field of road transport and for

interfaces with other modes of transport”, Available at

“http://ec.europa.eu/transport/themes/its/road/action_plan/,” Aug 6, 2010, pp L

207/1- 207/13.

2. Xinping Yan, Hui Zhang and Chaozhong Wu, “Research and Development of

Intelligent Transportation Systems,” in 11th International Symposium on

Distributed Computing and Applications to Business, Engineering & Science, Oct

19-22, 2012, pp 321-327.

3. National ITS Architecture Version 7.0, “http://www.its.dot.gov/arch/in

dex.htm,” July 10, 2012.

4. Tariq Samad and Anuradha Annaswamy, “Vehicle-to-Vehicle/Vehicle-to-

Infrastructure Control From The Impact of Control Technology,” IEEE Journals

and Magazines, www.ieeecss.org, 2011, Vol. 31, pp 26-27.

5. National Highway Traffic Safety Administration Report, “Vehicle-to-Vehicle

Communications: Readiness of V2V Technology for Application,” Available at

http://www.nhtsa.gov/About+NHTSA/Press+Releases/NHTSA-issues-advanced-

notice-of-proposed-rulemaking-on-V2V-communications, August 2014.

6. http://www.loc.gov/rr/scitech/mysteries/global.html, June 22, 2011

7. Robert Bosch, “CAN Specification Version 2.0,” Available at http://www.bosch-

semiconductors.de/media/pdf_1/canliteratur/can2spec.pdf

63
8. Federal Communications Commission, FCC 99-305. “FCC Report and Order,”

Available at

https://transition.fcc.gov/Bureaus/Engineering_Technology/News_Releases/1999/

nret9006.html,” October 1999.

9. Federal Communications Commission, FCC 03-024. “FCC Report and Order,”

Available at “https://info.publicintelligence.net/JSC-03-024.pdf,” February 2004.

10. B. Cash, “IEEE briefing on 5.9 GHz concepts and band plan,” Presentation in

DSRC standardization monthly meeting, 2002.

11. H. Krishnan and C. Kellum, “Use of communication in vehicle safety

application,” Internal Report of General Motors Company, 2002.

12. Jean Walrand and Pravin Varaiya, “High-performance Communications

Networks,” By Morgan Kaufmann, 2nd edition, 2000.

13. http://oceanservice.noaa.gov/facts/lidar.html, May 29, 2015

14. Wang Xing, Huiyan Chen and Huarong Ding, “The Application of Controller

Area Network on Vehicle,” in Vehicle Electronics Conference, 1999. (IVEC '99)

Proceedings of the IEEE International, Vol. 1, pp 455-458.

15. Sewon Lee, Bong-Su Cho, Yong-June Choi and Kwang-Ryul Baek,

“Implementation of MOST/CAN Network Protocol,” in Electrical and

Controlling Engineering (ICECE), International Conference on 16-18 Sept. 2011,

pp 5974-5977.

16. http://www.can-cia.org/, Aug 20, 2015.

17. IS0 11898 Vehicle-Digital Information Interchange-High Speed Communication

Controller Area Network (Version l), 1993.1 1.

64
18. IS0 Open System Interconnect (OSI) Model, IS0 7498.

19. http://www.obdii.com/connector.html, Aug 20, 2015.

20. Renjun Li, Chu Liu and Feng Luo, “A design for Automotive CAN Bus

Monitoring System,” in IEEE Vehicle Power and Propulsion Conference (VPPC),

September 3-5, 2008, Harbin, China, pp 1-5.

21. PDIUSBD12 Product data. Rev. 08.Philips, 20 December 2001.

22. Ashwin Gumaste, Rahul Singhai and Anirudh Sahoo, “IntelliCarTS: Intelligent

Car Transportation System,” Proc. IEEE LANMAN, June 2007.

23. Ashwin Gumaste, Rahul Singhai and Anirudh Sahoo, “Intelligent Vehicular

Transportation System (InVeTraS),” in Australasian Telecommunication

Networks and Applications Conference, December 2nd – 5th 2007, Christchurch,

New Zealand, pp 58-63.

24. Xue Yang, Jie Liu, N.F. Vaidya and Feng Zhao, “A Vehicle-to-Vehicle

Communication Protocol for Cooperative Collision Warning,” in Proceedings of

the First Annual International Conference on Mobile and Ubiquitous Systems:

Networking and Services, IEEE, 2004, pp 114-123.

25. M.J.F. Ashrafi, S. Yousefi, H. Karimi, S.M. Hosseini, H. Rostami and H.R.

Ataeian, “Highway Chain Collision Avoidance using Inter-Vehicular

Communications,” in 3rd International Conference on Computer and Knowledge

Engineering, October 31 & November 1, 2013, Ferdowsi University of Mashhad,

pp 135-140.

26. Liang Li, Guangquan Lu, Yunpeng Wang and Daxin Tian, “A Rear-end Collision

Avoidance System of Connected Vehicles,” in IEEE 17th International

65
 Conference on Intelligent Transportation Systems (ITSC) October 8-11, 2014.

Qingdao, China, pp 63-68.

27. Jong-Min Park, Hyun-Chul Choi and Se-young Oh, “Real-Time Vehicle

Detection in Urban Traffic Using Adaboost,” Proc. Of IEEE/RSJ International

Conference on IROS, pp. 3598 –3603, 2010.

28. Paul Viola and Michael J. Jones, “Robust Real-Time Face Detection,”

International Journal of Computer Vision 57(2), 2004, pp 137–154.

29. A. Khammari, F. Nashashibi, Y. Abramson, Y and C. Laurgeau, "Vehicle

detection combining gradient analysis and AdaBoost classification," in Intelligent

Transportation Systems, 2005. Proceedings IEEE 2005, pp 66-71.

30. Gwang Yul Song, Ki Yong Lee, and Joon Woong Lee, “Vehicle detection by

edge-based candidate generation and appearance-based classification,” Intelligent

Vehicles Symposium, 2008 IEEE, June 2008, pp 428-433.

31. Guan-Yu Chen, Po-Chun Shen, Chao-Yi Cho, M.S. Vinay and Jiun-In Guo “A

Forward Collision Avoidance System Adopting Multi-feature Vehicle Detection,”

in Consumer Electronics - Taiwan (ICCE-TW), 2014 IEEE International

Conference on 26-28 May 2014, pp-125 – 126.

32. P. Lytrivis, G. Thomaidis, M. Tsogas and A. Amditis, Member, IEEE; “An

Advanced Cooperative Path Prediction Algorithm for Safety Applications in

Vehicular Networks,” in IEEE Transactions on Intelligent Transportation

Systems, Vol. 12, No. 3, September 2011, pp 669-679.

66
33. Simon Julier and Jeffrey K. Uhlmann, “A general method for approximating

nonlinear transformations of probability distributions,” Robot. Res. Group, Dept.

Eng. Sci., Univ. Oxford, Oxford, U.K., 1996.

34. Eric A. Wan and Rudolph Van der Merwe, “The unscented Kalman filter,” in

Kalman Filtering and Neural Networks, S. Haykin, Ed. New York: Wiley, 2001,

pp. 221–280.

35. M. Tsogas, A. Polychronopoulos and A. Amditis, “Unscented Kalman filter

design for curvilinear motion models suitable for automotive safety applications,”

in Proc. 8th Int. Conf. Inf. Fusion, Philadelphia, PA, July 25-28, 2005.

36. A.P. Dempster, “A generalization of Bayesian inference,” J. R. Stat. Soc., Ser. B,

vol. 30, no. 2, pp. 205–247, 1968.

37. Glenn Shafer, “A Mathematical Theory of Evidence,” Princeton, NJ: Princeton

Univ. Press, 1976.

38. http://www.ubuntu.com/, Aug 20, 2015.

39. R. Braden (ed.), “RFC 1122, Requirements for Internet Hosts – Communication

Layers,” October 1989.

40. R. Braden (ed.), “RFC 1123, Requirements for Internet Hosts – Application and

Support,” October 1989

41. http://www.axisnetworkcabling.com/voice-data-cabling-installers-company.html,

Aug 5, 2015.

42. http://www.aradasystems.com, Aug 5, 2015.

43. http://www.adafruit.com/products/960, Aug 10, 2015.

44. http://www.vector.com/vi_candb_en.html, Aug 10, 2015.

67
45. https://www.python.org/, Aug 10, 2015.

46. K. Yedukondalu, A.D. Sarma and V.S. Srinivas, "Estimation and mitigation of

GPS multipath interference using adaptive filtering," Progress In

Electromagnetics Research M, Vol. 21, pp. 133-148, 2011.

47. http://www-nrd.nhtsa.dot.gov/Pubs/812160.pdf, June 2015.

48. http://www.census.gov/compendia/statab/2012/tables/12s1103.pdf, Aug 16, 2015.

49. http://www-nrd.nhtsa.dot.gov/Pubs/812139.pdf, Aug 16, 2015.

50. http://www.tc.gc.ca/media/documents/roadsafety/cmvtcs2012_eng.pdf, Aug 16,

2015.

51. http://www.cs.umd.edu/class/sum2003/cmsc311/Notes/Data/endian.html, Aug 20,

2015.

68
 APPENDICES

Appendix A

Big Endian Algorithm

Input: Data bytes, RcvDataLen

Output: FinalDecodedValue

Receive CAN ID = Data byte shift by 4 bytes of total data bytes

FOR ByteCounter = 4 to RcvDataLen

BinaryValue = IntToBinaryString(Data[ByteCounter])

END FOR

LenBinaryValue = BinaryValue.length()

FOR i = 0 to LenBinaryValue -8; i += 8

Push Back in vector ByteStrings(substr of BinaryValue(i,8))

END FOR

StartBitCount = 0

FOR ByteNum = 0 to ByteStrings.size()

Byte = ByteStrings[ByteNum]

RevByte = string reverse(Byte)

FOR RevBitPos = 0 to not RevByte.size()

IF (StartBitCount == datastructure -> signal -> bit_start)

BitPos = 7 – RevBitPos

BitPos = 8 * ByteNum + BitPos

FinalString = SubString of BinaryValue (BitPos, datastructure -> signal -> length)

69
DecodedValue = BinaryToInt(FinalString)

FinalValue = (DecodedValue * datastructure -> signal -> scale) + (datastructure -> signal

-> offset)

ELSE

StartBitCount++

END IF

END FOR

END FOR

70
 Appendix B

Little Endian Algorithm

Input: Data bytes, RcvDataLen

Output: FinalDecodedValue

Receive CAN ID = Data byte shift by 4 bytes of total data bytes

FOR ByteCounter = 4 to RcvDataLen

BinaryValue = IntToBinaryString(Data[ByteCounter])

END FOR

LenBinaryValue = BinaryValue.length()

FOR i = 0 to LenBinaryValue -8; i += 8

Push Back in vector ByteStrings(substr of BinaryValue(i,8))

END FOR

StartBitCount = 0

FOR ByteNum = 0 to ByteStrings.size()

Byte = ByteStrings[ByteNum]

RevByte = string reverse(Byte)

FOR RevBitPos = 0 to not RevByte.size()

IF (StartBitCount == datastructure -> signal -> bit_start)

BitPos = 7 – RevBitPos

BitPos = 8 * ByteNum + BitPos

BitPos = BitPos – (datastructure -> signal -> bit_len - 1)

FinalString = SubString of BinaryValue (BitPos, datastructure -> signal -> length)

DecodedValue = BinaryToInt(FinalString)

71
FinalValue = (DecodedValue * datastructure -> signal -> scale) + (datastructure -> signal

-> offset)

ELSE

StartBitCount++

END IF

END FOR

END FOR

72
 VITA AUCTORIS

NAME: Gaurav Sood

PLACE OF BIRTH: Malerkotla, India

YEAR OF BIRTH: 1986

EDUCATION: Sita Grammar School, Malerkotla, India, 2002

D.A.V College, Chandigarh, India, 2004

Guru Nanak Dev University, B.Tech., Amritsar, India,

2008

University of Windsor, M.Sc., Windsor, ON, Canada,

2015

73