Distributed filter state estimation of topological random switching in WSN

Abstract

A distributed filter design method for WSN with sensor nonlinearity saturation and random switching of network topology are proposed in this paper. A mass of sensor nodes are deployed in the sensor network, the target device is sensed and measured, and transmitted to the distributed filter through the network. In the filtering network, the local estimator receives the measurement information from the sensor node. The estimation from the neighboring nodes via a random time-varying topology to complete the state estimation and trajectory tracking for the objective system. The Bernoulli binary distribution is used to describe the random saturation nonlinearity of the sensor network, and the inhomogeneous Markov chain is adopted to represent the random switching topologies. The sufficient conditions are given as distributed filters in the form of linear matrix inequalities method. In the end, the effectiveness of this design method is illustrated by a simulation example.