Adaptive event sensing in networks of autonomous mobile agents

TL;DR

This paper presents a new distributed control algorithm for autonomous mobile agents to adaptively sense and record events in a region, switching between random and gradient-based movement modes based on local density estimates.

Contribution

It introduces a novel execution mode-based control algorithm for mobile agents that adaptively switches between movement strategies to optimize event detection.

Findings

Algorithm effectively balances exploration and exploitation.

Simulation results demonstrate adaptability to varying event densities.

Agents improve detection efficiency through mode switching.

Abstract

Given a connected region in two-dimensional space where events of a certain kind occur according to a certain time-varying density, we consider the problem of setting up a network of autonomous mobile agents to detect the occurrence of those events and possibly record them in as effective a manner as possible. We assume that agents can communicate with one another wirelessly within a fixed communication radius, and moreover that initially no agent has any information regarding the event density. We introduce a new distributed algorithm for agent control based on the notion of an execution mode, which essentially lets each agent roam the target region either at random or following its local view of a density-dependent gradient. Agents can switch back and forth between the two modes, and the precise manner of such changes depends on the setting of various parameters that can be adjusted as a function of the application at hand. We provide simulation results on some synthetic applications especially designed to highlight the algorithm's behavior relative to the possible execution modes.