Learning to Communicate with Deep Multi-Agent Reinforcement Learning

TLDR

The study focuses on multi‑agent systems that must learn communication protocols to maximize shared utility in partially observable environments. The authors propose two learning approaches, RIAL and DIAL, to enable agents to learn communication protocols. RIAL employs deep Q‑learning, while DIAL leverages back‑propagation through noisy communication channels, allowing centralized learning with decentralized execution. Experiments demonstrate that end‑to‑end learning of communication protocols is achievable in complex, partially observable environments, and that the introduced environments and engineering innovations are essential for success.

Abstract

We consider the problem of multiple agents sensing and acting in environments with the goal of maximising their shared utility. In these environments, agents must learn communication protocols in order to share information that is needed to solve the tasks. By embracing deep neural networks, we are able to demonstrate end-to-end learning of protocols in complex environments inspired by communication riddles and multi-agent computer vision problems with partial observability. We propose two approaches for learning in these domains: Reinforced Inter-Agent Learning (RIAL) and Differentiable Inter-Agent Learning (DIAL). The former uses deep Q-learning, while the latter exploits the fact that, during learning, agents can backpropagate error derivatives through (noisy) communication channels. Hence, this approach uses centralised learning but decentralised execution. Our experiments introduce new environments for studying the learning of communication protocols and present a set of engineering innovations that are essential for success in these domains.