PhD Oral Comprehensive Examination by Muhammad Anwar Shahid:"Machine Learning for Cybersecurity in Vehicular Adhoc Networks (VANET) "

Wednesday, June 23, 2021 - 13:00 to 16:00


The School of Computer Science would like to present…     

PhD. Comprehensive Exam by:  Muhammad Anwar Shahid 

Date: Wednesday June 23, 2021 
Time: 1:00pm-4:00pm 
Passcode: If interested in attending this event, contact the graduate secretary at


Vehicles on the roads play an important role in our daily lives. Increasing number of vehicles on the roads is producing more problems for traffic management authorities, drivers, and other people. Some of these problems include collisions, congestion, air pollution and fuel consumption. According to a report published by Canadian Motor Vehicle Traffic Collision in 2018, there were around 150,000 injuries, 11.5 million hours spent on the road and 22 million litres of fuel drained each year in Canada only. In this era of boom in technologies where we are promoting smart cities, addressing above mentioned issues is critical for road traffic and vehicles. Intelligent Transportation System (ITS) has proposed Vehicular Adhoc Network (VANET) that allows safety and non-safety communications between vehicles. VANET is a special class of Mobile Adhoc Network (MANET) that is designed to enable vehicle communication. In general, VANET consists of On-Board Unit (OBU), Road-Side Unit (RSU) and Central Authorities like Certificate Authority, Misbehavior Authority and so on. Federal Communication Congress (FCC) introduced Dedicated Short Range Communication (DSRC) for VANET in 1999. The DSRC is specially designed for vehicle communication based on the IEEE standard 802.11p and 5.9 GHz frequency band. DSRC supports communication between vehicles and infrastructure within a maximum range of 1000m. For long range communication, FCC has recently proposed Cellular-V2X technology which supports vehicle-to-everything communication by using mobile communication standards such as 3G, 4G, LTE, 5G and beyond 5G. Vehicles in VANET share safety and non-safety information with each other through Basic Safety Messages (BSM). BSM is broadcasted by each vehicle periodically to share its status. BSM, in general, contains core elements including position, heading, speed, steering wheel angle, and vehicle size. VANET requires stable, secure and uninterrupted communication for a successful delivery of BSMs. Security is one of the biggest challenges in VANET because vehicular communication is vulnerable to various attacks. In VANET, security must ensure that exchanged messages are not altered or created by malicious attackers. Denial of Service (DoS), Replay, GPS Spoofing and Sybil attacks are some of the known attacks in VANET.  
Timely detection of these attacks can play a vital role in securing vehicular communication. In this research, we focus on applying Machine Learning (ML) and Deep Learning (DL) algorithms to detect security attacks. One of the limitations in using these algorithms in VANET is the publicly available datasets. Datasets for machine learning based solutions must foresee qualitative attributes including accuracy, completeness, reliability, relevance, and timeliness. VANET (Vehicular ad hoc network) has witnessed an extremely limited publicly available dataset in terms of in-vehicle and inter-vehicle communications. One of the reasons could be access to higher capacity and computational resources to generate these kinds of datasets. VANET generates massive amount of data having high dimensional features and attributes. In literature, VeReMi (Vehicular Reference Misbehavior) dataset is the well known publicly available dataset for inter vehicle communications as a benchmark to conduct research using machine learning approaches. Though, VeReMi dataset contains position falsification attack only. 
This thesis explores the effectiveness of using machine learning approaches for cybersecurity in VANET. We will conduct our research to 1) generate dataset to include more attacks like DoS, DDoS, Sybil and Replay, 2) use of plausible, ML and DL models to detect attacks, 3) detect attacks at local and global level. Proposed framework would revoke the malicious attackers and help in improving safety and security of the system. 
Keywords: VANET, ITS, DSRC, C-V2X, Basic Safety Message, Security Attacks, Machine Learning, Deep Learning, Intrusion Detection System (IDS). 

PhD Dissertation Committee: 

Internal Reader: Dr. Imran Ahmad 
Internal Reader: Dr. Saja Al-Mamoori 
External Reader: Dr. Ning Zhang (Dept. of Electrical Engineering 
Advisor: Dr. Arunita Jaekel 

PhD Comprehensive Exam Announcement     Vector Institute in Artificial Intelligence, artificial intelligence approved logo


 5113 Lambton Tower 401 Sunset Ave. Windsor ON, N9B 3P4 (519) 253-3000 Ext. 3716 (working remotely)