In this work, we analyze the performance of a simple majority-rule protocol solving a fundamental coordination problem in distributed systems - \emph{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 $\sqrt{n}$, then the SMP with only two communication rounds attains consensus on the majority opinion of the network, and if this difference grows faster than $\sqrt{n}$, then the SMP reaches consensus on the majority opinion of the network in a single round, with probability converging to $1$ exponentially fast as $n \rightarrow \infty$. We also provide some converse results, showing that these requirements are not only sufficient, but also necessary.
翻译:在这项工作中,我们分析了解决分布式系统中基本协调问题的简单多数规则协议的性能,即在出现概率性电文损失的情况下,解决分布式系统中基本协调问题的简单多数规则协议( \ emph{ binal oplication) 。 使用大规模、 完全连接的、 由2美元代理商组成的网络的概率分析,我们证明简单多数协议(SMP)只在三轮通信中达成共识,概率接近1美元,因为美元增长到无限。 此外,如果持有不同意见的代理商数量之间的差别以$/sqrt{n}的速率增长,那么只有两轮通信的SMP就网络多数意见达成了共识,如果这一差异增长快于$/sqrt{n},那么SMP在单轮中就网络多数意见达成共识,并有可能以1美元指数快速递增至1美元。 我们还提供了一些相反的结果,表明这些要求不仅足够,而且是必要的。