Learning Efficient Multi-agent Communication: An Information Bottleneck Approach

TL;DR

This paper introduces IMAC, a method for learning efficient multi-agent communication protocols under bandwidth constraints, using an information bottleneck approach to generate informative, low-entropy messages and effective scheduling.

Contribution

The paper proposes a novel IMAC method that jointly learns communication protocols and scheduling for multi-agent systems under bandwidth limitations, grounded in information theory.

Findings

IMAC converges faster than baseline methods.

IMAC achieves more efficient communication under limited bandwidth.

The approach is effective in various cooperative and competitive tasks.

Abstract

We consider the problem of the limited-bandwidth communication for multi-agent reinforcement learning, where agents cooperate with the assistance of a communication protocol and a scheduler. The protocol and scheduler jointly determine which agent is communicating what message and to whom. Under the limited bandwidth constraint, a communication protocol is required to generate informative messages. Meanwhile, an unnecessary communication connection should not be established because it occupies limited resources in vain. In this paper, we develop an Informative Multi-Agent Communication (IMAC) method to learn efficient communication protocols as well as scheduling. First, from the perspective of communication theory, we prove that the limited bandwidth constraint requires low-entropy messages throughout the transmission. Then inspired by the information bottleneck principle, we learn a valuable and compact communication protocol and a weight-based scheduler. To demonstrate the efficiency of our method, we conduct extensive experiments in various cooperative and competitive multi-agent tasks with different numbers of agents and different bandwidths. We show that IMAC converges faster and leads to efficient communication among agents under the limited bandwidth as compared to many baseline methods.