Distributed filtering for uncertain time-varying Markovian jump linear systems over binary sensor network

This paper deals with the scalable distributed filtering problem with binary measurements for a class of discrete time-varying uncertain Markovian jump linear systems. The system and the original measurement are subject to random noise with known means and variances. A Markov chain is introduced to formulate the switching of system parameters, the uncertainties of which are characterized by the norm-bounded constraint. The useful information for estimation purpose is extracted via the crossings between the original binary measurements and the desirable binary measurements. By using the information from neighboring nodes, sufficient conditions are established for every node to guarantee the existence of the distributed filter such that the filtering error satisfies the exponential boundedness in the mean square. The filter gains are calculated on each node by means of recursive linear matrix inequalities, which achieve the desirable scalability of distributed filtering owing to such a manner. Finally, a simulation example demonstrates the applicability of the distributed filtering scheme proposed in this paper.