Cybersecurity of connected and automated vehicles via traffic anomaly detection

Principal Investigator

  • Raphael Stern, Associate Professor, Civil, Environmental and Geo-Engineering

Summary

This report explores cyber vulnerabilities that exist in new approaches for actuated traffic flow management where individual connected and automated vehicles share information about their arrival to a reinforcement learning based traffic signal. Using this approach, potential cyberattacks include false data injection attacks, where vehicles report false data. We find that new signal techniques are vulnerable to such attacks, deteriorating intersection level of service, and also present approaches to overcome this vulnerability through federation.

Project Details

Research Reports