A Game-Theoretic Multi-Patroller Approach for Critical Infrastructure Monitoring

Critical infrastructure systems often include outdated or compromised devices that pose significant security risks. Ensuring the integrity of such devices is challenging due to the high cost and operational impact of constant monitoring. In this paper, we address this challenge by formulating it as a patrolling security game with multiple defensive agents and a single attacker. We propose a revisited algorithm that coordinates multiple agents to optimally patrol and verify the device integrity, effectively extending patrolling security games to a multi-patroller setting. The algorithm is validated through simulation experiments, laying the groundwork for systems capable of improving attack detection rates and optimizing resource usage compared to baseline approaches. The results highlight the practical value of our multi-agent patrolling strategy in enhancing the cybersecurity of critical infrastructures.