Traffic Engineering in Large-scale Networks with Generalizable Graph Neural Networks

2025-03-31

AI Summary

Traditional traffic engineering (TE) algorithms suffer from poor scalability in ultra-large-scale networks (e.g., cloud WANs, LEO satellite constellations), while existing learning-based methods exhibit weak generalization, high training overhead, and difficulty in enforcing capacity constraints. Method: We propose a novel “learning-to-optimize” TE paradigm—learning the *algorithm* for computing optimal solutions, rather than predicting solutions directly. Our approach employs a lightweight graph neural network with topology-agnostic feature encoding and an end-to-end algorithmic learning framework, ensuring zero-shot generalization across unseen topologies and traffic patterns while strictly satisfying link capacity constraints and solution feasibility. Contributions/Results: On 5,000-node networks, our method achieves optimality gaps <3%; maintains effectiveness when tested on networks up to 20× larger than those seen during training; accelerates inference by 84% over exact solvers; and reduces training and inference latency by 2–4 orders of magnitude versus state-of-the-art learning-based TE methods.

Technology Category

• Search and Optimization: Learning to Search
• Machine Learning: Optimization
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web

Abstract

Traffic engineering (TE) in large-scale computer networks has become a fundamental yet challenging problem, owing to the swift growth of global-scale cloud wide-area networks or backbone low-Earth-orbit satellite constellations. To address the scalability issue of traditional TE algorithms, learning-based approaches have been proposed, showing potential of significant efficiency improvement over state-of-the-art methods. Nevertheless, the intrinsic limitations of existing learning-based methods hinder their practical application: they are not generalizable across diverse topologies and network conditions, incur excessive training overhead, and do not respect link capacities by default. This paper proposes TELGEN, a novel TE algorithm that learns to solve TE problems efficiently in large-scale networks, while achieving superior generalizability across diverse network conditions. TELGEN is based on the novel idea of transforming the problem of"predicting the optimal TE solution"into"predicting the optimal TE algorithm", which enables TELGEN to learn and efficiently approximate the end-to-end solving process of classical optimal TE algorithms. The learned algorithm is agnostic to the exact network topology or traffic patterns, and can efficiently solve TE problems given arbitrary inputs and generalize well to unseen topologies and demands. We trained and evaluated TELGEN on random and real-world networks with up to 5000 nodes and 106 links. TELGEN achieved less than 3% optimality gap while ensuring feasibility in all cases, even when the test network had up to 20x more nodes than the largest in training. It also saved up to 84% solving time than classical optimal solver, and could reduce training time per epoch and solving time by 2-4 orders of magnitude than latest learning algorithms on the largest networks.

Problem

• Addressing scalability in large-scale network traffic engineering
• Improving generalizability across diverse topologies and conditions
• Reducing training overhead while respecting link capacities

Innovation

• Generalizable Graph Neural Networks for TE
• Transforms TE solution prediction to algorithm prediction
• Efficiently approximates classical optimal TE algorithms