Position-Aware Communication via Self-Attention for Multi-Agent Reinforcement Learning | IEEE Conference Publication | IEEE Xplore