Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Li, Xinran, Wang, Xiaolu, Bai, Chenjia, Zhang, Jun
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912249943162880
author Li, Xinran
Wang, Xiaolu
Bai, Chenjia
Zhang, Jun
author_facet Li, Xinran
Wang, Xiaolu
Bai, Chenjia
Zhang, Jun
contents In cooperative multi-agent reinforcement learning (MARL), well-designed communication protocols can effectively facilitate consensus among agents, thereby enhancing task performance. Moreover, in large-scale multi-agent systems commonly found in real-world applications, effective communication plays an even more critical role due to the escalated challenge of partial observability compared to smaller-scale setups. In this work, we endeavor to develop a scalable communication protocol for MARL. Unlike previous methods that focus on selecting optimal pairwise communication links-a task that becomes increasingly complex as the number of agents grows-we adopt a global perspective on communication topology design. Specifically, we propose utilizing the exponential topology to enable rapid information dissemination among agents by leveraging its small-diameter and small-size properties. This approach leads to a scalable communication protocol, named ExpoComm. To fully unlock the potential of exponential graphs as communication topologies, we employ memory-based message processors and auxiliary tasks to ground messages, ensuring that they reflect global information and benefit decision-making. Extensive experiments on large-scale cooperative benchmarks, including MAgent and Infrastructure Management Planning, demonstrate the superior performance and robust zero-shot transferability of ExpoComm compared to existing communication strategies. The code is publicly available at https://github.com/LXXXXR/ExpoComm.
format Preprint
id arxiv_https___arxiv_org_abs_2502_19717
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning
Li, Xinran
Wang, Xiaolu
Bai, Chenjia
Zhang, Jun
Multiagent Systems
Artificial Intelligence
Machine Learning
In cooperative multi-agent reinforcement learning (MARL), well-designed communication protocols can effectively facilitate consensus among agents, thereby enhancing task performance. Moreover, in large-scale multi-agent systems commonly found in real-world applications, effective communication plays an even more critical role due to the escalated challenge of partial observability compared to smaller-scale setups. In this work, we endeavor to develop a scalable communication protocol for MARL. Unlike previous methods that focus on selecting optimal pairwise communication links-a task that becomes increasingly complex as the number of agents grows-we adopt a global perspective on communication topology design. Specifically, we propose utilizing the exponential topology to enable rapid information dissemination among agents by leveraging its small-diameter and small-size properties. This approach leads to a scalable communication protocol, named ExpoComm. To fully unlock the potential of exponential graphs as communication topologies, we employ memory-based message processors and auxiliary tasks to ground messages, ensuring that they reflect global information and benefit decision-making. Extensive experiments on large-scale cooperative benchmarks, including MAgent and Infrastructure Management Planning, demonstrate the superior performance and robust zero-shot transferability of ExpoComm compared to existing communication strategies. The code is publicly available at https://github.com/LXXXXR/ExpoComm.
title Exponential Topology-enabled Scalable Communication in Multi-agent Reinforcement Learning
topic Multiagent Systems
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2502.19717