Event-based Distributed State Estimation over a WSN with False Data Injection Attack

Abstract: In this paper, we consider the security issue in a wireless sensor network (WSN). We investigate distributed state estimation over a WSN under hostile attacks. To defend against malicious attacks, we propose a novel event-based mechanism, with which each sensor decides to send data to its neighboring sensors according to a data detection condition. When the detection condition is satisfied, the sensor is triggered at time step k, then it estimates the target state and sends its observation data to its neighbors; Otherwise, the sensor can only receive the data from its neighboring sensors. We provide a sufficient condition to guarantee the stability of the proposed estimator under the malicious attacks, and a numerical example to illustrate the effectiveness of the proposed event-based estimator.