City Research Online

Abstract

In recent years, intelligent transport systems have become an area of interest, as they can make transport safer and more convenient for passengers. This has been attempted in public transport before, but never in private forms of transport such as cars, which are far less predictable, requiring the user to relinquish existing control over their vehicle and its decision making to another, non-human system.

It is known that vehicle accidents are very costly and happen frequently, with human error being one of the key reasons behind this. Therefore, a better decision-making system that does not make such error is of great interest, which is where the concept of Connected and Autonomous Vehicles (CAVs) comes into play.

Currently, there are either connected vehicles or autonomous vehicles, which are vehicles that communicate with one another and ones that can make their own decisions respectively, but not both. CAVs aim to combine these, using the connected aspect to increase information about surroundings (i.e., obstacles, traffic, and vehicle manoeuvres) in combination with the autonomous features to allow a vehicle to make accurate, safe, and informed decisions.

However, the systems of CAVs are not perfect, being susceptible to faults and attacks which can have severe safety implications. Many of these attacks are done by exploiting the newfound connectivity of the vehicle, through which they can manipulate information being sent to the vehicle to affect the decision making of the autonomous component, sometimes even controlling the vehicle itself.

To assess how well a vehicle holds up against attacks, I must look at cyber-resilience, which is what I aim to assess in this project, with proper, realistic models and results to back up any arguments made regarding resilience. Stochastic models of the vehicle operation under attacks will be created in the Mobius software, looking at how various parts of the system behave under attack and in the presence of different defence mechanisms.

For this research, the focus of attacks will be on those affecting the safety of the vehicle, i.e., attacks which disrupt the function of the vehicle significantly, which lead to accidents that would not have happened without attacks. By the end of this research, a working model of CAVs under attack will be offered, showing their resilience to safety-compromising attacks and the effect of employing different defences.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA76 Computer software T Technology > TL Motor vehicles. Aeronautics. Astronautics
Departments:	School of Science & Technology > Department of Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads