RouteNet: Leveraging Graph Neural Networks for network modeling and optimization in SDN

TL;DR

RouteNet employs Graph Neural Networks to accurately predict network performance metrics like delay and loss in SDN, enabling better network management and optimization across diverse topologies and traffic conditions.

Contribution

This paper introduces RouteNet, a novel GNN-based model that generalizes network KPI predictions to arbitrary topologies, routing, and traffic, improving over existing models.

Findings

RouteNet predicts delay and loss with a worst-case MRE of 15.4%.

The model generalizes well to unseen topologies and traffic conditions.

KPI predictions enable efficient routing and network planning.

Abstract

Network modeling is a key enabler to achieve efficient network operation in future self-driving Software-Defined Networks. However, we still lack functional network models able to produce accurate predictions of Key Performance Indicators (KPI) such as delay, jitter or loss at limited cost. In this paper we propose RouteNet, a novel network model based on Graph Neural Network (GNN) that is able to understand the complex relationship between topology, routing, and input traffic to produce accurate estimates of the per-source/destination per-packet delay distribution and loss. RouteNet leverages the ability of GNNs to learn and model graph-structured information and as a result, our model is able to generalize over arbitrary topologies, routing schemes and traffic intensity. In our evaluation, we show that RouteNet is able to predict accurately the delay distribution (mean delay and jitter) and loss even in topologies, routing and traffic unseen in the training (worst case MRE=15.4%). Also, we present several use cases where we leverage the KPI predictions of our GNN model to achieve efficient routing optimization and network planning.