Set-membership estimator design with privacy-preserving for sensor networks: A state-decomposition-based approach

Set-membership estimation over open sensor networks is a current research hotspot. It remains a risk of data leakage under highly open network channels. This paper focuses on establishing set-membership estimator design strategy under a privacy-preserving framework. Firstly, with the aid of state decomposition methods, a new group of set-membership estimators with privacy-preserving are originally established. Wherein the real measurement is divided into two parts for information interaction and performance compensation respectively. The proposed estimator structure contributes to the security of private data by confusing eavesdroppers. Secondly, a sufficient condition is established to guarantee that the ellipsoidal estimation sets are updated properly. Compared with the existing results on set-membership estimation, the developed one is less conservative. Thirdly, a rigorous theoretical analysis is proposed to justify that the proposed estimator is sufficiently robust to confuse eavesdroppers. Lastly, a simulation example is given to evaluate the estimation and privacy-preserving performance of the proposed technique.