Agent-Based Modeling for Analysis of Cyber Attacks on the Intelligent Transportation System

The advanced computing and communication between components within the intelligent transportation system creates vulnerabilities that can be exploited by bad actors. To ensure the safety of autonomous vehicles and secure the infrastructure, we must first understand how a cyber attack propagates through the intelligent transportation system. This propagation occurs through communications between individual agents and entire subsystems. By understanding the impact of component and subsystem interactions, we can better fortify the intelligent transportation system against cyber attacks. This paper presents an agent-based model for detection of and recovery from cyber attacks on the intelligent transportation system for autonomous vehicles. In this model, vehicles, pedestrians, and intelligent infrastructure are represented as distinct agent populations. Simulations were run to test the spread of cyber attack while considering the origin and aggressiveness of the attack, and detection and defense capabilities. The simulations showed that clusters of distinct agents will form based on communication ranges. These clusters can quickly merge as new agents are introduced. Preventing the introduction of new agents can isolate subsystems affected by an attack.