Decoupling Communication from Policy: Robust MARL under Bandwidth Constraints

TL;DR

This paper introduces a novel MARL architecture that decouples communication from policy, enabling robust coordination under bandwidth constraints with minimal performance loss.

Contribution

The paper proposes SLIM, a minimal architecture that separates communication from policy representation, and introduces a normalized bandwidth budget to improve multi-agent reinforcement learning under communication limits.

Findings

Achieves state-of-the-art performance on MARL benchmarks.

Maintains robustness with minimal performance degradation under bandwidth reduction.

Scales effectively with limited communication bandwidth.

Abstract

Communication enables coordination in multi-agent reinforcement learning (MARL), but many real-world applications, e.g., search-and-rescue with drone swarms, operate under severe bandwidth constraints. Many communication architectures still expose a coupled bottleneck in which a shared latent representation is used for both policy execution and inter-agent communication. Consequently, reducing message size directly limits the policy's latent space, often leading to significant performance degradation. We address this with two contributions. First, we introduce $\beta$, a normalised per-agent bandwidth budget that unifies sparsity, rounds, and message dimension into a single comparable constraint. Second, we provide SLIM, a minimal architecture that decouples the communication pathway from the policy's latent representation, allowing us to isolate the effect of bandwidth from the effect of policy capacity while benefiting from in-step communication. We evaluate our method on several partially-observable MARL benchmarks, where communication is essential. Our approach achieves state-of-the-art performance and exhibits scalability and robustness under limited communication, with only marginal degradation as bandwidth is reduced.