Event-Triggered Distributed Cubature Kalman Filtering Algorithm With Stealthy Attacks Over Sensor Networks

This article investigates the security problem of distributed state estimation for nonlinear systems subject to stealthy attacks and limited energy. First, a novel detection strategy for a nonlinear information consensus filter is designed to resist the stealthy adversary which can modify the data transmitted through the wireless network. Unlike existing attack detection strategies, the proposed defense strategy is capable of simultaneously verifying the authenticity of the received local estimate and error covariance. Afterward, considering the limited communication resources, an event-triggered distributed cubature Kalman filtering algorithm with the aforementioned detection strategy is presented to fuse the local information. This algorithm can reduce communication consumptions and guarantee good estimation precision for sensor networks with stealthy attacks and limited energy. Subsequently, the stability properties of the developed nonlinear filtering algorithm are presented. Finally, two examples are given to demonstrate the effectiveness of the proposed approach.