Multi-Agent Common Knowledge Reinforcement Learning

TL;DR

This paper introduces MACKRL, a hierarchical reinforcement learning algorithm that leverages common knowledge among agents to enable complex decentralized coordination in multi-agent systems, improving performance on challenging tasks.

Contribution

The paper proposes a novel stochastic actor-critic algorithm, MACKRL, which learns hierarchical policies conditioned on common knowledge for decentralized multi-agent coordination.

Findings

MACKRL outperforms existing methods on complex coordination tasks.

The hierarchical policy structure effectively utilizes common knowledge.

MACKRL generalizes to fully decentralized policies as a special case.

Abstract

Cooperative multi-agent reinforcement learning often requires decentralised policies, which severely limit the agents' ability to coordinate their behaviour. In this paper, we show that common knowledge between agents allows for complex decentralised coordination. Common knowledge arises naturally in a large number of decentralised cooperative multi-agent tasks, for example, when agents can reconstruct parts of each others' observations. Since agents an independently agree on their common knowledge, they can execute complex coordinated policies that condition on this knowledge in a fully decentralised fashion. We propose multi-agent common knowledge reinforcement learning (MACKRL), a novel stochastic actor-critic algorithm that learns a hierarchical policy tree. Higher levels in the hierarchy coordinate groups of agents by conditioning on their common knowledge, or delegate to lower levels with smaller subgroups but potentially richer common knowledge. The entire policy tree can be executed in a fully decentralised fashion. As the lowest policy tree level consists of independent policies for each agent, MACKRL reduces to independently learnt decentralised policies as a special case. We demonstrate that our method can exploit common knowledge for superior performance on complex decentralised coordination tasks, including a stochastic matrix game and challenging problems in StarCraft II unit micromanagement.