Air-Ground Coordination Communication by Multi-Agent Deep Reinforcement Learning | IEEE Conference Publication | IEEE Xplore