Globally optimal distributed information filter for mobile sensor network

This paper studies the networked estimation of the state of a linear dynamical system. The focus is on developing a distributed estimation algorithm which could achieve the same performance as the centralized ones. A communication protocol which formally describes the network topology is introduced. By performing this protocol, the sum of the information vector contributions is computed, which is then used to update the local estimations through utilizing the information filter equations. One attractive feature of this algorithm is that it only takes finite communication times to derive globally optimal estimations. Monte Carlo simulations show that the proposed algorithm outperforms the other existing distributed estimation algorithm with the same communication burden.