Abstract: Multi-agent reinforcement learning (MARL) has become a fundamental component of next-generation wireless communication systems. Theoretically, although MARL has the advantages of low ...