Networked Communication for Decentralised Agents in Mean-Field Games

TL;DR

This paper introduces a networked communication framework for mean-field games with decentralized agents, improving scalability and practical convergence, and demonstrating empirical advantages over existing methods.

Contribution

It extends mean-field game algorithms with networked communication, providing theoretical guarantees and practical enhancements for decentralized multi-agent systems.

Findings

Accelerates learning compared to independent agents

Performs similarly to centralized algorithms in practice

Improves robustness to update failures and population changes

Abstract

Methods like multi-agent reinforcement learning struggle to scale with growing population size. Mean-field games (MFGs) are a game-theoretic approach that can circumvent this by finding a solution for an abstract infinite population, which can then be used as an approximate solution for the $N$-agent problem. However, classical mean-field algorithms usually only work under restrictive conditions. We take steps to address this by introducing networked communication to MFGs, in particular to settings that use a single, non-episodic run of $N$ decentralised agents to simulate the infinite population, as is likely to be most reasonable in real-world deployments. We prove that our architecture's sample guarantees lie between those of earlier theoretical algorithms for the centralised- and independent-learning architectures, varying dependent on network structure and the number of communication rounds. However, the sample guarantees of the three theoretical algorithms do not actually result in practical convergence times. We thus contribute practical enhancements to all three algorithms allowing us to present their first empirical demonstrations. We then show that in practical settings where the theoretical hyperparameters are not observed, giving fewer loops but poorer estimation of the Q-function, our communication scheme still respects the earlier theoretical analysis: it considerably accelerates learning over the independent case, which hardly seems to learn at all, and often performs similarly to the centralised case, while removing the restrictive assumption of the latter. We provide ablations and additional studies showing that our networked approach also has advantages over both alternatives in terms of robustness to update failures and to changes in population size.