Event-Triggered Resilient Estimator for Nonlinear Cyber-Physical Systems Under DoS Attacks

This paper tackles the challenge of state estimation in nonlinear cyber-physical systems under Denial-of-Service (DoS) attacks. To address this, it proposes a novel event-triggered resilient estimator, offering enhanced resistance against such disruptions. Firstly, to account for DoS attacks, a random attack model is established. This model paves the way for a dynamic event-triggered mechanism, which reduces network load by strategically decreasing the transmission rate of measurement data. Next, this paper tackles the challenge of recursively solving coupled Riccati-like difference equations. Through this approach, an estimation gain is obtained that minimizes the upper bound (UB) on the estimation error covariance. Leveraging random analytic theory, the paper rigorously proves that the estimation error is probabilistically bounded. This theoretical foundation is then validated by a example showcasing the proposed estimator's performance.