Virtual Clustering for Distributed Consensus-Based Estimation in Cooperative Networks

This paper presents a new approach for distributed consensus-based estimation with low communication effort in cooperative networks, where a group of nodes cooperates to estimate source messages with information exchange achieving consensus among all nodes. A new distributed estimation algorithm is developed by adopting the novel approach of virtual clustering, in which the size of exchanged data is reduced during the distributed processing by the partial transmission after data clustering. Moreover, the required communication effort and estimation performance of the algorithm are evaluated for networks with different topology and varying system parameters. We show that communication cost is reduced considerably by the proposed algorithm while keeping the estimation performance.