Continuous-Time Distributed Dynamic Programming for Networked Multi-Agent Markov Decision Processes

TL;DR

This paper develops and analyzes continuous-time distributed dynamic programming algorithms for multi-agent Markov decision processes, enabling agents to learn optimal policies through local rewards and neighbor communication.

Contribution

It introduces novel distributed DP algorithms inspired by optimization methods and proves their convergence, advancing multi-agent reinforcement learning.

Findings

Convergence of the proposed algorithms is rigorously proven.

The algorithms facilitate decentralized learning in multi-agent systems.

The framework enables future development of distributed temporal difference learning.

Abstract

The main goal of this paper is to investigate continuous-time distributed dynamic programming (DP) algorithms for networked multi-agent Markov decision problems (MAMDPs). In our study, we adopt a distributed multi-agent framework where individual agents have access only to their own rewards, lacking insights into the rewards of other agents. Moreover, each agent has the ability to share its parameters with neighboring agents through a communication network, represented by a graph. We first introduce a novel distributed DP, inspired by the distributed optimization method of Wang and Elia. Next, a new distributed DP is introduced through a decoupling process. The convergence of the DP algorithms is proved through systems and control perspectives. The study in this paper sets the stage for new distributed temporal different learning algorithms.