Cluster-Based Fast Distributed Consensus

In this paper, we propose cluster-based distributed averaging algorithms in forms of fixed iteration and random gossiping. Nodes within a cluster maintain the same value via broadcasting by the cluster-head, and information exchange occurs between neighboring clusters. Clustering essentially allows nodes in neighboring clusters to be joined, hence the resultant graph is well-connected and the algorithm converges much faster. Moreover, since the number of clusters is much smaller than the number of nodes, the communication and computation burden of the consensus algorithm is significantly reduced.