Energy Absorbable Optimal DoS Attack Strategy for Remote State Estimation in CPSs With Two-Hop Networks

This paper focuses on the investigation of optimal attack power strategies in cyber-physical systems with a two-hop network subject to two malicious attackers. The objective is to analyze the effects of DoS attacks on wireless communication channels, specifically targeting the degradation of the system’s estimation performance. The attackers can exploit energy from the surrounding environment and employ it to interfere with the communication of information from the sensor to the remote estimator. Both attackers share the same goal of maximizing the error covariance at the remote estimator during the transmission process. To address this problem, we formulate it as a multi-agent markov decision process (MAMDP) and proposed a Q-learning algorithm to determine the optimal attack power strategies. The work provides insights into effective strategies for countering the attackers’ disruptive actions and mitigating the impact on the system’s estimation performance. Finally, simulation examples are offered to validate and demonstrate the theoretical results.