Enhancing Heterogeneous Multi-Agent Cooperation in Decentralized MARL via GNN-driven Intrinsic Rewards

TL;DR

This paper introduces CoHet, a GNN-based intrinsic motivation method that enhances cooperation among heterogeneous agents in decentralized multi-agent reinforcement learning, especially under partial observability and reward sparsity.

Contribution

We propose a novel GNN-driven intrinsic motivation approach for decentralized heterogeneous multi-agent RL, addressing limitations of prior methods that rely on centralized training and agent indexing.

Findings

CoHet outperforms state-of-the-art methods in multiple benchmarks.

The intrinsic motivation improves cooperation under reward sparsity.

The approach is robust to increasing heterogeneity and agent numbers.

Abstract

Multi-agent Reinforcement Learning (MARL) is emerging as a key framework for various sequential decision-making and control tasks. Unlike their single-agent counterparts, multi-agent systems necessitate successful cooperation among the agents. The deployment of these systems in real-world scenarios often requires decentralized training, a diverse set of agents, and learning from infrequent environmental reward signals. These challenges become more pronounced under partial observability and the lack of prior knowledge about agent heterogeneity. While notable studies use intrinsic motivation (IM) to address reward sparsity or cooperation in decentralized settings, those dealing with heterogeneity typically assume centralized training, parameter sharing, and agent indexing. To overcome these limitations, we propose the CoHet algorithm, which utilizes a novel Graph Neural Network (GNN) based intrinsic motivation to facilitate the learning of heterogeneous agent policies in decentralized settings, under the challenges of partial observability and reward sparsity. Evaluation of CoHet in the Multi-agent Particle Environment (MPE) and Vectorized Multi-Agent Simulator (VMAS) benchmarks demonstrates superior performance compared to the state-of-the-art in a range of cooperative multi-agent scenarios. Our research is supplemented by an analysis of the impact of the agent dynamics model on the intrinsic motivation module, insights into the performance of different CoHet variants, and its robustness to an increasing number of heterogeneous agents.