Optimal Trajectory Learning for UAV-BS Video Provisioning System: A Deep Reinforcement Learning Approach

The unmanned aerial vehicle (UAV) based data transmission is highlighted for next-generation communication system by both academia and industry. The UAV, which is dynamically associated with mobile users, can take a role of base station (BS) as service provider (SP), for various types of scenarios. For this sake, it is important that the UAV-BS should be hovered in the air with obeying optimal trajectory for minimizing delay, which is caused by enormous computation of data transmission, and the trajectory can be controlled by centralized macro base station (MBS). In this paper, we propose deep reinforcement learning approach for computing optimal trajectories of distributed UAV-BS with low-latency overhead to enable efficient UAV communication in next generation wireless system.