Resilient State Estimate for Maneuvering Target Tracking under Cyber-Attacks

In this article, we address distributed state estimate problem for targets with uncertain maneuvers, where each sensor in the network is vulnerable to cyber-attacks. For calculating the accurate local estimate results, target maneuvers dynamic model is given by introducing adaptive jerk component. Specifically, by considering the average frequency and duration, denial-of-service (DoS) attack model has been established. With DoS model and distributed estimate framework, the resilient state estimation algorithm is designed by local filter, resilient, and consensus fusion stages. The final distributed estimates of targets can be calculated by each sensor node repeatedly exchanging information contribution with its neighbors in the consensus fusion stage. Moreover, with some mild conditions, the bounds of the information matrix and estimate error are proved to be existed. Numerical simulation for state estimate of targets with unknown maneuvers is performed to verify the performance of the designed algorithm.