ACCNet: Actor-Coordinator-Critic Net for "Learning-to-Communicate" with Deep Multi-agent Reinforcement Learning
This addresses the challenge of scalable and adaptive communication for multi-agent reinforcement learning, though it appears incremental as it builds on existing actor-critic and deep learning technologies.
The paper tackles the problem of learning communication protocols in multi-agent systems under partially observable environments, proposing the ACCNet framework which achieves better results than several baselines in both continuous and discrete action spaces.
Communication is a critical factor for the big multi-agent world to stay organized and productive. Typically, most previous multi-agent "learning-to-communicate" studies try to predefine the communication protocols or use technologies such as tabular reinforcement learning and evolutionary algorithm, which can not generalize to changing environment or large collection of agents. In this paper, we propose an Actor-Coordinator-Critic Net (ACCNet) framework for solving "learning-to-communicate" problem. The ACCNet naturally combines the powerful actor-critic reinforcement learning technology with deep learning technology. It can efficiently learn the communication protocols even from scratch under partially observable environment. We demonstrate that the ACCNet can achieve better results than several baselines under both continuous and discrete action space environments. We also analyse the learned protocols and discuss some design considerations.