Robust Filtering Algorithm for Uncertain Systems with Observation Losses in Sensor Network

In this paper, robust minimum variance filtering problem is considered for discrete time-varying systems with observation losses. The system is subjected to time-varying norm-bounded parameter uncertainties in both the state and output matrices, and the observation losses are described by a Bernoulli process with a known probability. Based on an upper bound on the variance of the state estimation error, a robust filter is derived by minimizing the prescribed upper bound in the sense of the matrix norm. Eventually, an algorithm suitable for online computation is summarized and a simulation example is presented to demonstrate the effectiveness of the proposed algorithms.