Event-based distributed recursive filtering for state-saturated systems with redundant channels

Abstract

In this paper, the event-based distributed recursive filtering problem is investigated for a class of discrete-time state-saturated systems subject to random occurring nonlinearities and measurement losses over the wireless sensor network. In the addressed measurement model, the sensors are assumed to have redundant communication channels that are helpful in increasing the probability of successfully delivering the measurements. At each intelligent node, the local estimation is obtained based on its own measurement and those transmitted from its neighbors according to the sensor topology. In order to reduce the bandwidth consumption and estimator update frequencies, an event-based signal transmission strategy is employed as opposed to the traditional time-based one. An upper bound for the estimation error covariance is constructed at each time step, which is shown to be the solution of a Riccati-like difference equation. Subsequently, the estimator parameter is designed to minimize such an upper bound. Moreover, the performance of the proposed estimator is analyzed by discussing how the packet losses of the measurements affect the obtained upper bound of the error covariance. Finally, a numerical simulation is exploited to show the effectiveness of the proposed distributed filter design algorithm.