Decentralized Control for Multi-agent Missions Based on Flocking Rules

This paper addresses the problem of having a multi-agent system converging to multiple dynamic rendezvous areas for networks of agents with no localization sensors and limited communication capabilities. The localization is performed by measurement/communication towers that determine the noisy position and velocity of each relevant agent and transmit them using a directional antenna. There is no assumption on the connectedness of the network topology, which is unknown to all agents. The proposed solution consists of an improved flocking-based movement algorithm tailored to the proposed scenario, with mechanisms to prevent collisions. The performance of the solution is presented through multiple simulations for a multitude of environments. By an appropriate selection of an utility function, these empirical results hint to the possibility of realizing different types of missions intended for multi-agent systems using swarm behavior.