Distributed Fusion for Multisensor Systems With Censored Measurements Under Round-Robin Protocol

In this article, distributed fusion is investigated for a multisensor system affected by censored measurements and packet dropout. A Round-Robin protocol is proposed to alleviate communication burden and avoid packet collisions in communication networks with packet dropout. A local estimator with the Round-Robin protocol is designed to acquire local estimates for the multisensor system under censored measurements and packet dropout. A distributed fusion algorithm is introduced to enhance estimation accuracy using a fast covariance intersection fusion (FCIF) method. Sufficient conditions on the designed local estimator are given to guarantee the boundedness of estimation error covariance matrix. A numerical example and an experiment are given to demonstrate effectiveness of the proposed algorithm.