Information-weighted consensus filtering for cyber-physical systems under stealthy attacks

This paper investigates the distributed state estimation problem in wireless sensor networks that employ amplify-and-forward relays, particularly in the presence of stealthy attacks. By extending the Kalman consensus filter, a consensus protocol on weighted directed graphs is introduced to develop a novel distributed minimum mean square error estimator. The proposed method efficiently integrates local measurements and relay information while mitigating the impact of malicious interventions. Assuming resource-constrained adversaries, a weighted consensus strategy is adopted to preserve directed information flow under stealthy attacks, thereby minimizing the estimation error. Furthermore, a sufficient condition is derived to ensure the boundedness of the estimation error in the presence of attacks. Finally, simulation results are provided to demonstrate the effectiveness and resilience of the proposed framework.