Communication Efficiency And Security For Multi-Agent Reinforcement Learning