Distributed Reinforcement Learning for Flexible and Efficient UAV Swarm Control

Over the past few years, use of swarms Unmanned Aerial Vehicles (UAVs) in monitoring and remote area surveillance applications has become widespread thanks to price reduction increased capabilities drones. The drones swarm need cooperatively explore an unknown area, order identify monitor interesting targets, while minimizing their movements. In this work, we propose a distributed Reinforcement Learning (RL) approach that scales larger without modifications. proposed framework relies on possibility for UAVs exchange some information through communication channel, achieve context-awareness implicitly coordinate swarm's actions. Our experiments show method can yield effective strategies, which are robust channel impairments, easily deal with non-uniform distributions targets obstacles. Moreover, when agents trained specific scenario, they adapt new one minimal additional training. We also our achieves better performance compared computationally intensive look-ahead heuristic.