Decentralized information-theoretic task assignment for searching and tracking of moving targets

Abstract

This paper presents a framework for information-theoretic based task assignment of multiple UAVs for searching and tracking moving targets. A stochastic approach which uses probability density functions is introduced as the information gathering metric. This algorithm yields trajectory that minimizes the uncertainty of unknown information. The presented task assignment is based on the negotiation activated when one of the defined conditions is satisfied. This task assignment can generate admissible solutions with minimal computational time. Cost computed by each agent from its information gathering layer is used in the task assignment layer to allocate the priority of tasks for each agent in a decentralized architecture. Our algorithm is validated on a search and track scenario with three fixed-wing UAVs and twelve moving targets.