Class
Article
Graduation Year
2018
College
College of Engineering
Department
Electrical and Computer Engineering Department
Faculty Mentor
Prof. Chris Winstead
Presentation Type
Oral Presentation
Abstract
Human mistakes are the main source of fatal accidents and daily traffic congestion. Recent researches have focused on assisting drivers to mitigate traffic fatalities and create more enjoyable drive experiences. Vehicle platooning or cooperative adaptive cruise control (CACC) has been stated as one of the most effective solutions to tackle this problem. Platooning concept involves a group of vehicles act as a single unit through coordination of movements. This concept draws a special attention among academia and governmental and non-governmental organizations. Recently, there have been several demonstrations, which have introduced the potential benefits of this idea, e.g. safety enhancement, increase roadway capacity, and improve traffic efficiency. While many aspects of platooning such as transportation impacts, mechanical and control concerns are still under investigation, very limited amount of work has studied platooning in an adversarial environment. To design safe distributed controllers and networks, it is essential to understand the possible attacks that can be applied against platoons.
In this work, we design a set of insider attack and abnormal behaviors that are implemented in a car platoon. For example, an attack has been introduced, in which attacker exploits the platoon controller to cause collisions and disrupt the performance of platoon. In the small platoon, this successful attack can be carried through a malicious member of platoon solely by changing its motion (acceleration and deceleration) and gains of the controller. However, the attack should be executed through a collaborative effort in large platoons. In this case, the main attacker not only adjusts its gains and motion to accomplish the attack but also collaborates with the other attacker. Another attacker only uses the gain modification technique. The strong point of the proposed attack is that attacker stays intact while in existing works from literature, the attacker gets affected during an attack.
Location
Room 155
Start Date
4-13-2017 10:30 AM
End Date
4-13-2017 11:45 AM
Cybersecurity of Autonomous Vehicle Platooning
Room 155
Human mistakes are the main source of fatal accidents and daily traffic congestion. Recent researches have focused on assisting drivers to mitigate traffic fatalities and create more enjoyable drive experiences. Vehicle platooning or cooperative adaptive cruise control (CACC) has been stated as one of the most effective solutions to tackle this problem. Platooning concept involves a group of vehicles act as a single unit through coordination of movements. This concept draws a special attention among academia and governmental and non-governmental organizations. Recently, there have been several demonstrations, which have introduced the potential benefits of this idea, e.g. safety enhancement, increase roadway capacity, and improve traffic efficiency. While many aspects of platooning such as transportation impacts, mechanical and control concerns are still under investigation, very limited amount of work has studied platooning in an adversarial environment. To design safe distributed controllers and networks, it is essential to understand the possible attacks that can be applied against platoons.
In this work, we design a set of insider attack and abnormal behaviors that are implemented in a car platoon. For example, an attack has been introduced, in which attacker exploits the platoon controller to cause collisions and disrupt the performance of platoon. In the small platoon, this successful attack can be carried through a malicious member of platoon solely by changing its motion (acceleration and deceleration) and gains of the controller. However, the attack should be executed through a collaborative effort in large platoons. In this case, the main attacker not only adjusts its gains and motion to accomplish the attack but also collaborates with the other attacker. Another attacker only uses the gain modification technique. The strong point of the proposed attack is that attacker stays intact while in existing works from literature, the attacker gets affected during an attack.