GDDR: GNN-based Data-Driven Routing

TL;DR

This paper investigates using Graph Neural Networks combined with Deep Reinforcement Learning for data-driven routing in networks, demonstrating comparable performance to existing methods and better generalization across different network topologies.

Contribution

It introduces a GNN-based policy architecture for network routing that generalizes well across topologies, advancing the application of GNNs in systems and network optimization.

Findings

GNN-based approach performs at least as well as MLP-based methods.

GNNs enable generalization to different network topologies without retraining.

The method has potential applications beyond routing in systems research.

Abstract

We explore the feasibility of combining Graph Neural Network-based policy architectures with Deep Reinforcement Learning as an approach to problems in systems. This fits particularly well with operations on networks, which naturally take the form of graphs. As a case study, we take the idea of data-driven routing in intradomain traffic engineering, whereby the routing of data in a network can be managed taking into account the data itself. The particular subproblem which we examine is minimising link congestion in networks using knowledge of historic traffic flows. We show through experiments that an approach using Graph Neural Networks (GNNs) performs at least as well as previous work using Multilayer Perceptron architectures. GNNs have the added benefit that they allow for the generalisation of trained agents to different network topologies with no extra work. Furthermore, we believe that this technique is applicable to a far wider selection of problems in systems research.