A Multi-Agent, Policy-Gradient approach to Network Routing

2025/12/02 by Nigel Tao, Tao, Nigel, Jonathan Baxter +3
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #Network Traffic and Congestion Control #Networking and Internet Architecture (cs.NI) #Peer-to-Peer Network Technologies

paper · pdf · doi:10.48550/arxiv.2512.03211

openalex created_date 2016/06/24 · openalex publication_date 2025/12/02 · openalex updated_date 2026/07/28

Abstract

Network routing is a distributed decision problem which naturally admits numerical performance measures, such as the average time for a packet to travel from source to destination. OLPOMDP, a policy-gradient reinforcement learning algorithm, was successfully applied to simulated network routing under a number of network models. Multiple distributed agents (routers) learned co-operative behavior without explicit inter-agent communication, and they avoided behavior which was individually desirable, but detrimental to the group's overall performance. Furthermore, shaping the reward signal by explicitly penalizing certain patterns of sub-optimal behavior was found to dramatically improve the convergence rate.

Citations

• Avoiding Braess' Paradox through Collective Intelligence
• Learning Finite-State Controllers for Partially Observable Environments

Cited by

• Modelling the Dynamic Joint Policy of Teammates with Attention\n Multi-agent DDPG
• MAMRL: Exploiting Multi-agent Meta Reinforcement Learning in WAN Traffic Engineering

Related

• Packet Routing with Graph Attention Multi-agent Reinforcement Learning
• DRAMA: A Dynamic Packet Routing Algorithm using Multi-Agent Reinforcement Learning with Emergent Communication
• Navigating Communication Networks with Deep Reinforcement Learning
• Navigating Communication Networks with Deep Reinforcement Learning
• Relational Deep Reinforcement Learning for Routing in Wireless Networks