Cellular UAV-to-Device Communications: Trajectory Design and Mode Selection by Multi-Agent Deep Reinforcement Learning | IEEE Journals & Magazine | IEEE Xplore