Cooperative Multiagent Deep Reinforcement Learning for Reliable Surveillance via Autonomous Multi-UAV Control

CCTV-based surveillance using unmanned aerial vehicles (UAVs) is considered a key technology for security in smart city environments. This paper creates case where the UAVs with CCTV-cameras fly over area flexible and reliable services. should be deployed to cover large while minimize overlapping shadow areas system. However, operation of subject high uncertainty, necessitating autonomous recovery systems. work develops multi-agent deep reinforcement learning-based management scheme industry applications. The core idea this employs autonomously replenishing UAV's deficient network requirements communications. Via intensive simulations, our proposed algorithm outperforms state-of-the-art algorithms terms coverage, user support capability, computational costs.