Reducing Communication Entropy in Multi-Agent Reinforcement Learning

Lebin Yu, Yunbo Qiu, Qiexiang Wang, Xu-Dong Zhang, Jian Wang

Keywords: multi-agent reinforcement learning, multi-agent communication, low entropy

Abstract: Communication in multi-agent reinforcement learning has been drawing attention recently for its significant role in cooperation. However, multi-agent systems may suffer from limitations on communication resource and thus need efficient communication techniques in real-world scenarios. According to the Shannon-Hartley theorem, messages to be transmitted reliably in worse channels requires lower entropy. Therefore, we aim to reduce message entropy in multi-agent communication. A fundamental challenge in this is that the gradients of entropy are either 0 or infinity, disabling gradient-based methods. To handle it, we propose a pseudo gradient descent scheme, which reduces entropy by adjusting the distributions of messages wisely. We conduct experiments on six environment settings and two base communication frameworks and find that our scheme can reduce communication entropy by up to 90% with nearly no loss of performance.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Reinforcement Learning (eg, decision and control, planning, hierarchical RL, robotics)