Connected and Autonomous Vehicle Cybersecurity
Mentor 1
Lingfeng Wang
Location
Union Wisconsin Room
Start Date
27-4-2018 1:00 PM
Description
Modern vehicles are implementing features of autonomy at a staggering rate. Many models already include assisted parking and automatic braking, supported by the coordination of several on-board sensors. As vehicles transition toward full autonomy and driverless vehicles become an everyday reality, a robust network of connectivity must also be implemented. A secure Vehicular Sensor Network (VSN) for exchanging data enables enhanced road safety features and anti-congestion algorithms for a smooth flow of traffic. Along with these advances in technology, however, more complex cyber vulnerabilities continue to emerge. Researchers have already highlighted a wide variety of security flaws in modern systems, allowing attackers to steal or modify information, deny services, and even hijack control of moving vehicles. These studies have inspired a recent emphasis on the development of defensive techniques. The in-vehicle network consists of many Electronic Control Units (ECUs) which exchange messages over a Controller Area Network (CAN) bus system. For connected networks of vehicles, the Ad hoc On-demand Distance Vector (AODV) protocol is emerging as the primary means of communication. Focusing mostly on these two communication methods, the aim of this study is twofold. Firstly, to compile and describe known attacks against modern and emerging vehicle technology. Second is to propose state of the art defense mechanisms against attacks in this realm. Proposed techniques include message verification and malicious node detection schemes for the CAN protocol within vehicles, as well as AODV communication between vehicles.
Connected and Autonomous Vehicle Cybersecurity
Union Wisconsin Room
Modern vehicles are implementing features of autonomy at a staggering rate. Many models already include assisted parking and automatic braking, supported by the coordination of several on-board sensors. As vehicles transition toward full autonomy and driverless vehicles become an everyday reality, a robust network of connectivity must also be implemented. A secure Vehicular Sensor Network (VSN) for exchanging data enables enhanced road safety features and anti-congestion algorithms for a smooth flow of traffic. Along with these advances in technology, however, more complex cyber vulnerabilities continue to emerge. Researchers have already highlighted a wide variety of security flaws in modern systems, allowing attackers to steal or modify information, deny services, and even hijack control of moving vehicles. These studies have inspired a recent emphasis on the development of defensive techniques. The in-vehicle network consists of many Electronic Control Units (ECUs) which exchange messages over a Controller Area Network (CAN) bus system. For connected networks of vehicles, the Ad hoc On-demand Distance Vector (AODV) protocol is emerging as the primary means of communication. Focusing mostly on these two communication methods, the aim of this study is twofold. Firstly, to compile and describe known attacks against modern and emerging vehicle technology. Second is to propose state of the art defense mechanisms against attacks in this realm. Proposed techniques include message verification and malicious node detection schemes for the CAN protocol within vehicles, as well as AODV communication between vehicles.
