Distributed fusion estimation for sensor networks with Markov-based nonuniform samplings and multiplicative noise

The distributed fusion estimation problem is studied for multiple sensors networks with multiplicative noise and nonuniform samplings, where the nonuniform sampling process is modeled as a multi-state Markov chain. Based on the state augmented method, a multi-sensor single-rate system with Markovian jumping parameters is obtained. Further, local estimators are derived and a distributed covariance intersection fusion estimator is proposed, which avoids the computation of the estimation error cross-covanriance matrices between any two local estimators, thereby reducing the computaional cost.