Decoupling Communication from Policy: Robust MARL under Bandwidth Constraints
For MARL practitioners in bandwidth-limited domains (e.g., drone swarms), this work provides a principled way to handle communication constraints without sacrificing policy capacity.
The paper addresses bandwidth constraints in multi-agent reinforcement learning (MARL) by introducing a normalized bandwidth budget and a decoupled communication architecture (SLIM). The approach achieves state-of-the-art performance with only marginal degradation as bandwidth is reduced.
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 $β$, 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.