Simple Majority Consensus in Networks with Unreliable Communication

Abstract

In this work, we analyze the performance of a simple majority-rule protocol solving a fundamental coordination problem in distributed systems—binary majority consensus—in the presence of probabilistic message loss. Using probabilistic analysis for a large-scale, fully-connected, network of 2n agents, we prove that the Simple Majority Protocol (SMP) reaches consensus in only three communication rounds, with probability approaching 1 as n grows to infinity. Moreover, if the difference between the numbers of agents that hold different opinions grows at a rate of n, then the SMP with only two communication rounds attains consensus on the majority opinion of the network, and if this difference grows faster than n, then the SMP reaches consensus on the majority opinion of the network in a single round, with probability converging to 1 as exponentially fast as n→∞. We also provide some converse results, showing that these requirements are not only sufficient, but also necessary.