Distributed Consensus in Networks

Abstract

Distributed algorithms have gained a lot of attention during recent years. Their application in industry, particularly in wireless sensor networks has motivated researchers to try to design them in order to be less resource-consuming (e.g. memory and power), faster, and more reliable. There have been numerous distributed algorithms for different types of problems in the context of distributed algorithms. We are interested in a fundamental coordination problem namely the majority consensus problem. In the majority consensus problem nodes try to find the opinion of the majority in a network of interest. As our first contribution and motivated by the distributed binary consensus problem in [1] we propose a distributed algorithm for multivalued consensus in complete graphs. As our second contribution we propose an algorithm for the optimization of the binary interval consensus algorithm pioneered by Ben ezit et al in [2]. Finally we use binary interval consensus algorithm to design a framework for error-free consensus in dynamic networks using which nodes can leave or join the network during or after the consensus process.