Searching Algorithm for Autonomous Distributed Agents based on Chemotactic Behavior of C. elegans

In this paper, the authors proposed a searching algorithm for autonomous distributed agents based on the behavior of C. elegans, a small nematode. As C. elegans exhibits chemotaxis despite its small neural system, it can be a model of an autonomous distributed searching agent. They modified the existing chemotactic behavior model of C. elegans so that a modeled agent behaves like a “successful” individual of C. elegans that visits a global peak with high probability, without being captured by local peaks. They translated the modified behavior model to a simplified motion algorithm of an autonomous agent, and examined the search performance through comparisons with other algorithms. They also applied a Gaussian mixture model to trajectories exhibited by agents driven by the proposed algorithm and obtained a potential map to locate a global peak.