Fusion Estimation for Stochastic Systems with Mixed Attacks and Correlated Noises

The information fusion estimation problems are investigated for a class of stochastic systems with mixed network attacks and correlated noises in this paper. The system and observation noises are correlated at the same time stamps. The signal encountered mixed attacks including deception attacks and denial of service (DoS) attacks during the transmission from sensors to remote estimators. The optimal centralized fusion estimators including filter, predictor and smoother are proposed in the linear unbiased minimum variance sense. Furthermore, based on local filters and estimation error cross-covariance matrices between any two local filters, the distributed fusion filter weighted by matrix is presented, which has better robustness and flexibility. Simulation example shows the effectiveness of the proposed algorithms.