In-Progress: Reinforcement Learning for Cyberattack Defense in Autonomous Intersection Management Systems

This study proposes an RL-enabled framework to enhance cybersecurity in Autonomous Intersection Management (AIM) systems. While AIM improves traffic flow and safety, its network dependence makes it vulnerable to cyber threats. Using Proximal Policy Optimization (PPO) and Advantage Actor-Critic (A2C), the framework detects and mitigates DoS, data manipulation, and malware attacks in a digital-twin simulation. Results demonstrate RL's potential for adaptive, resilient security in intelligent transportation systems.