Adaptive Graph Coordination Strategy in Multiagent Reinforcement Learning

doi:10.1109/TG.2025.3629681

Adaptive Graph Coordination Strategy in Multiagent Reinforcement Learning

2026-03-01

0

PRE

AI

邢登鹏 (Xing, Dengpeng) *

DOI：10.1109/TG.2025.3629681

Original

Original request for help

Save

Abstract

En 中文

Many real-world applications involve a team of agents who must coordinate each other's policies in real-time to achieve a shared goal. Previous studies mainly focus on decentralized control to maximize common rewards, with little consideration of coordination between control policies, which is critical in dynamic and complicated environments. Viewing this issue, we propose a novel adaptive graph coordination strategy that factorizes the joint policy into an adaptive graph generator and a graph-based coordinated policy. We employ a difference-aware module to control when to generate graphs and an encoder-decoder module to acquire the underlying decision graph structure. Moreover, we introduce DAGness- and DAG depth-constrained optimization to adjust the graph structure and strike a balance between efficiency and performance. We also present a graph-based coordinated policy to make asynchronous decisions based on the interagent coordination dependencies implied in the generated graph. Empirical evaluations on some cooperative multiagent environments demonstrate the superiority of the proposed method, with faster convergence and more efficient coordinated policies.

Keywords:

Generators

Decision making

Training

Optimization

Reinforcement learning

Multi-agent systems

Games

Probabilistic logic

Feature extraction

Directed acyclic graph

Action coordination graph (ACG)

directed acyclic graph (DAG)

multiagent systems (MASs)

reinforcement learning

Adaptive Graph Coordination Strategy in Multiagent Reinforcement Learning

Abstract

Journal

Organization