Consensus‐based unscented Kalman filtering over sensor networks with communication protocols

Abstract This article is concerned with the consensus‐based distributed filtering problem for a class of general nonlinear systems over sensor networks with communication protocols. In order to avoid data collisions, the stochastic protocol and the Round‐Robin protocol are respectively introduced to schedule the data transmission between each node and its neighboring ones. A consensus‐based unscented Kalman filtering (UKF) algorithm is developed for the purpose of estimating the system states over sensor networks subject to communication protocols. Moreover, the exponential boundedness of estimation error in mean square is proved for the proposed algorithm. Finally, compared with the extended Kalman filtering, an experimental simulation example is provided to validate the effectiveness of the consensus‐based UKF algorithm.