TMAC: a Transformer-based partially observable multi-agent communication method

Effective communication plays a crucial role in coordinating the actions of multiple agents. Within the realm of multi-agent reinforcement learning, agents have the ability to share information with one another through communication channels, leading to enhanced learning outcomes and successful goal...

Full description

Saved in:
Bibliographic Details
Main Authors: Xuesi Li, Shuai Xue, Ziming He, Haobin Shi
Format: Article
Language:English
Published: PeerJ Inc. 2025-04-01
Series:PeerJ Computer Science
Subjects:
Online Access:https://peerj.com/articles/cs-2758.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849733009318608896
author Xuesi Li
Shuai Xue
Ziming He
Haobin Shi
author_facet Xuesi Li
Shuai Xue
Ziming He
Haobin Shi
author_sort Xuesi Li
collection DOAJ
description Effective communication plays a crucial role in coordinating the actions of multiple agents. Within the realm of multi-agent reinforcement learning, agents have the ability to share information with one another through communication channels, leading to enhanced learning outcomes and successful goal attainment. Agents are limited by their observations and communication ranges due to increasingly complex location arrangements, making multi-agent collaboration based on communication increasingly difficult. In this article, for multi-agent communication in some partially observable scenarios, we propose a Transformer-based Partially Observable Multi-Agent Communication algorithm (TMAC), which improves agents extracting features and generating output messages. Meanwhile, a self-message fusing module is proposed to obtain features from multiple sources. Therefore, agents can achieve better collaboration through communication. At the same time, we performed experimental verification in the surviving and the StarCraft Multi-Agent Challenge (SMAC) environments where agents had limited local observation and could only communicate with neighboring agents. In two test environments, our method achieves an improvement in performance 6% and 10% over the baseline algorithm, respectively. Our code is available at https://gitee.com/xs-lion/tmac.
format Article
id doaj-art-df3e29c9839d4c328d3167176fde8f38
institution DOAJ
issn 2376-5992
language English
publishDate 2025-04-01
publisher PeerJ Inc.
record_format Article
series PeerJ Computer Science
spelling doaj-art-df3e29c9839d4c328d3167176fde8f382025-08-20T03:08:09ZengPeerJ Inc.PeerJ Computer Science2376-59922025-04-0111e275810.7717/peerj-cs.2758TMAC: a Transformer-based partially observable multi-agent communication methodXuesi LiShuai XueZiming HeHaobin ShiEffective communication plays a crucial role in coordinating the actions of multiple agents. Within the realm of multi-agent reinforcement learning, agents have the ability to share information with one another through communication channels, leading to enhanced learning outcomes and successful goal attainment. Agents are limited by their observations and communication ranges due to increasingly complex location arrangements, making multi-agent collaboration based on communication increasingly difficult. In this article, for multi-agent communication in some partially observable scenarios, we propose a Transformer-based Partially Observable Multi-Agent Communication algorithm (TMAC), which improves agents extracting features and generating output messages. Meanwhile, a self-message fusing module is proposed to obtain features from multiple sources. Therefore, agents can achieve better collaboration through communication. At the same time, we performed experimental verification in the surviving and the StarCraft Multi-Agent Challenge (SMAC) environments where agents had limited local observation and could only communicate with neighboring agents. In two test environments, our method achieves an improvement in performance 6% and 10% over the baseline algorithm, respectively. Our code is available at https://gitee.com/xs-lion/tmac.https://peerj.com/articles/cs-2758.pdfMulti-agent reinforcement learningCommunicationAttention mechanism
spellingShingle Xuesi Li
Shuai Xue
Ziming He
Haobin Shi
TMAC: a Transformer-based partially observable multi-agent communication method
PeerJ Computer Science
Multi-agent reinforcement learning
Communication
Attention mechanism
title TMAC: a Transformer-based partially observable multi-agent communication method
title_full TMAC: a Transformer-based partially observable multi-agent communication method
title_fullStr TMAC: a Transformer-based partially observable multi-agent communication method
title_full_unstemmed TMAC: a Transformer-based partially observable multi-agent communication method
title_short TMAC: a Transformer-based partially observable multi-agent communication method
title_sort tmac a transformer based partially observable multi agent communication method
topic Multi-agent reinforcement learning
Communication
Attention mechanism
url https://peerj.com/articles/cs-2758.pdf
work_keys_str_mv AT xuesili tmacatransformerbasedpartiallyobservablemultiagentcommunicationmethod
AT shuaixue tmacatransformerbasedpartiallyobservablemultiagentcommunicationmethod
AT ziminghe tmacatransformerbasedpartiallyobservablemultiagentcommunicationmethod
AT haobinshi tmacatransformerbasedpartiallyobservablemultiagentcommunicationmethod