In this work, we consider distributed agreement tasks in microbial distributed systems under stochastic population dynamics and competitive interactions. We examine how competitive exclusion can be used to solve distributed agreement tasks in the microbial setting. To this end, we develop a new technique for analyzing the time to reach competitive exclusion in systems with several competing species under biologically realistic population dynamics. We use this technique to analyze a protocol that exploits competitive interactions to solve approximate majority consensus efficiently in synthetic microbial systems. We show that direct competition dynamics reach majority consensus with high probability when the initial gap between the species is small, i.e., $\Omega(\sqrt{n \log n})$, where $n$ is the initial population size of the majority species. In contrast, we show that indirect competition alone is not efficient: for example, solving majority consensus with high probability requires an initial gap of $\Omega(n)$. To corroborate our analytical results, we use computer simulations to show that these consensus dynamics occur within practical time scales.
翻译:在这项工作中,我们考虑在随机种群动态和竞争性互动下在微生物分布系统中分配分配协议任务。我们研究如何利用竞争性排斥解决微生物环境中分配协议任务。为此,我们开发了一种新的技术,用于分析时间,以便在生物现实的种群动态下,与几个相互竞争的物种在系统中实现竞争性排斥。我们利用这一技术分析一项议定书,利用竞争性互动,在合成微生物系统中有效解决大致多数人的共识。我们显示,当物种之间的初始差距很小时,直接竞争动态就极有可能达成多数共识,即,美元(sqrt{n\log n}$),而美元是大多数物种的初始人口规模。相反,我们显示,单靠间接竞争是不有效率的:例如,在极有可能的情况下,实现多数人的共识,就需要美元(n)美元的初步差距。为了证实我们的分析结果,我们利用计算机模拟来显示这些共识动态是在实际时间范围内发生的。