Interference-Aware K-Step Reachable Communication in Multi-Agent Reinforcement Learning

Ziyu Cheng; Jinsheng Ren; Zhouxian Jiang; Chenzhihang Li; Rongye Shi; Bin Liang; Jun Yang

cs.AIMar 16, 2026

Interference-Aware K-Step Reachable Communication in Multi-Agent Reinforcement Learning

Authors: Ziyu Cheng, Jinsheng Ren, Zhouxian Jiang, Chenzhihang Li, Rongye Shi, Bin Liang, Jun Yang

Organizations: School of Software, Beihang University · School of Artificial Intelligence, Beihang University

Abstract

Effective communication is pivotal for addressing complex collaborative tasks in multi-agent reinforcement learning (MARL). Yet, limited communication bandwidth and dynamic, intricate environmental topologies present significant challenges in identifying high-value communication partners. Agents must consequently select collaborators under uncertainty, lacking a priori knowledge of which partners can deliver task-critical information. To this end, we propose Interference-Aware K-Step Reachable Communication (IA-KRC), a novel framework that enhances cooperation via two core components: (1) a K-Step reachability protocol that confines message passing to physically accessible neighbors, and (2) an interference-prediction module that optimizes partner choice by minimizing interference while maximizing utility. Compared to existing methods, IA-KRC enables substantially more persistent and efficient cooperation despite environmental interference. Comprehensive evaluations confirm that IA-KRC achieves superior performance compared to state-of-the-art baselines, while demonstrating enhanced robustness and scalability in complex topological and highly dynamic multi-agent scenarios.

Figures & tables

Figure 1: Limitations of common neighborhood constraints in MARL communication. (a) Euclidean distance may misrepresent actual accessibility: agents A and B appear close (blue dashed line) but are separated by a long traversable path (yellow arrow); (b) Vision-based constraints capture physical proximity better but miss reachable agents hidden from view; (c) Even with direct visibility, hostile interference (red region) can block cooperation, forcing detours (yellow arrows).

Figure 2: Overview of IA-KRC framework for MARL communication. The K - step Reachability Module restricts communication to agents within a physically reachable domain based on shortest transition distance. The Interference Prediction Module evaluates potential conflicts or adversarial effects, selecting low-interference partners. Combined, these modules enable dynamic grouping and robust leader–follower collaboration in complex environments.

Panel of Figure 3 — Figure 3: Illustration of the time-varying regulation (a) and the multi-layer map structure (b).

Method	Dense-Obstacle Map			Maze-Structure Map
	FW	HW/S	FL	FW	HW/S	FL
DPP	88.37 $\pm$ 3.12	88.37 $\pm$ 3.45 / 0.87M	2.8 $\pm$ 1.3	69.63 $\pm$ 4.28	69.63 $\pm$ 4.10 / 2.00M	15.2 $\pm$ 1.2
Euclid	78.81 $\pm$ 2.76	83.51 $\pm$ 2.94 / 0.61M	2.7 $\pm$ 1.2	77.89 $\pm$ 3.31	77.89 $\pm$ 3.05 / 1.99M	6.3 $\pm$ 1.3
SOG(Vision)	83.51 $\pm$ 2.67	83.51 $\pm$ 2.43 / 1.99M	2.6 $\pm$ 1.1	72.73 $\pm$ 3.88	72.73 $\pm$ 3.40 / 1.99M	15.0 $\pm$ 1.1
SOG(RL-Vision)	81.11 $\pm$ 3.01	83.60 $\pm$ 2.58 / 0.49M	7.5 $\pm$ 1.4	62.68 $\pm$ 4.90	62.68 $\pm$ 4.21 / 2.00M	3.8 $\pm$ 1.1
CommFormer	79.37 $\pm$ 2.25	79.37 $\pm$ 2.77 / 1.99M	7.6 $\pm$ 1.5	70.40 $\pm$ 3.56	70.40 $\pm$ 3.12 / 1.99M	15.1 $\pm$ 1.2

Table 1: Final cumulative win rate (FW), highest cumulative win rate (HW) within 2.0M steps for Dense-Obstacle and Maze-Structure maps. Additionally, we report final failure rate (FL) corresponding to FW.

Panel of Figure 7 — Figure 7: Group structure visualization: (a), (b), and (c) correspond to different algorithms; nodes with same color in same group; star=leader; red=isolated.

Table 7

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 8: Transition cost map computed from the start agent (yellow) after multi-layer map and interference map processing. Blue regions indicate low traversal cost due to high ally density and minimal enemy interference, while yellow regions represent high traversal cost caused by distant locations and strong adversarial agent interference.

Parameter	Value	Parameter	Value	Parameter	Value
General RL Parameters
Learning Rate	5e-4	Optimizer	Adam	Discount Factor ( $\gamma$ )	0.99
Batch Size	32	Replay Buffer Size	5000	Target Update Interval	200 ep.
Epsilon Start	1.0	Epsilon Finish	0.05	Epsilon Anneal Time	500k steps
Experience Buffer Size	10000	Opponent Learning Rate	1e-3
IA-KRC Framework and Interference Prediction

Appendix

Table 4: Key Hyperparameter Settings

Multi-Agent Communication

Multi-agent communication studies how autonomous agents exchange information to coordinate decisions, share knowledge, and act together under constraints.

See all 375 papers

Robust and Efficient Communication for Multi-Agent Learning
Sep 14, 2026Rafael Pina, Varuna De Silva, Corentin ArtaudSequential MARL Multi-Agent Communication
Institute for Digital Technologies · Loughborough University London, United Kingdom
LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning
May 18, 2026Sangjun Bae, Yisak Park, Sanghyeon Lee +1LLM-Guided RL Multi-Agent Systems
Graduate School of Artificial Intelligence, UNIST, Ulsan, South Korea.
Learning Multi-Agent Communication Protocol: Study on Information Entropy Efficiency in MARL
Jun 5, 2026Xinren Zhang, Zixin Zhong, Jiadong YuSequential MARL Agent Communication Protocols
The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China

Sep 14, 2026J/K move · Enter open · S save

Robust and Efficient Communication for Multi-Agent Learning

Rafael Pina, Varuna De Silva, Corentin Artaud

Institute for Digital Technologies · Loughborough University London, United Kingdom

Sequential MARL Multi-Agent Communication Cooperative MARL Multi-Agent Reinforcement Learning

Open paper AskPDF arXiv

Effective communication is a cornerstone of distributed intelligence in Multi-Agent Reinforcement Learning (MARL), yet ensuring that generated messages are both informative and robust to physical constraints remains a significant challenge. This paper introduces Multi-Agent Regularized Communication (MARC), a novel framework inspired by information-theoretic principles of conditional mutual information. MARC employs an attention-based architecture coupled with a unique message regularization mechanism designed to minimize uncertainty regarding future system states, thereby inducing the learning of highly representative communication protocols. Crucially, we evaluate MARC under stringent communication bottlenecks and lossy channels, simulating the real-world constraints of autonomous robotic networks and decentralized systems. Our results demonstrate that MARC significantly outperforms state-of-the-art methods in complex cooperative domains. Furthermore, we provide a deep analysis of message characteristics, proving that MARC maintains high operational performance even under significant data compression, offering a scalable path for deploying intelligent agents in resource-constrained environments.

Abstract

Multi-Agent Communication

Explore similar work

Robust and Efficient Communication for Multi-Agent Learning

LLM-Guided Communication for Cooperative Multi-Agent Reinforcement Learning

Learning Multi-Agent Communication Protocol: Study on Information Entropy Efficiency in MARL