A Cybersecurity Testbed for Connected and Autonomous Vehicle Systems
Primary Investigator:
Satish Ukkusuri
Ruichen Tan, Shiveswarran Ratneswaran, Zengxiang Lei
Abstract
This project seeks to build a sophisticated physical testbed designed to evaluate the real-world network-level impact of cyberattacks on CAV fleets. This project extends the co-simulation framework developed by our lab, which integrated METS-R and CARLA, by bringing simulation closer to reality. Specifically, the physical testbed will feature a detailed environment mapped similarly to CARLA's simulated maps, allowing us to project cyberattacks modeled in the simulation directly onto the physical testbed. This enables us to study how attacks observed in the simulation manifest in real-world scenarios, including their impact on sensor fusion, V2X communication, and vehicle behavior, such as car-following dynamics. The testbed will also allow us to examine network-level cascading effects, such as congestion caused by compromised vehicles, providing insights into the broader system impacts. Ultimately, this project aims to bridge the gap between simulation and reality, creating a robust platform for advancing cybersecurity research and improving the resilience of future autonomous vehicle systems with more controllable settings and low cost.