From GNNs to Symbolic Surrogates via Kolmogorov-Arnold Networks for Delay Prediction

2025-12-23

AI Summary

To address the challenge of accurate flow latency prediction in modern communication networks, this paper proposes a three-stage modeling paradigm: “heterogeneous graph neural network → interpretable symbolic surrogate model.” We innovatively integrate Kolmogorov–Arnold Networks (KAN) into message passing and attention mechanisms to construct KAMP-Attn, a heterogeneous GNN tailored for network topology and traffic dynamics. Furthermore, we introduce a graph-structure-preserving closed-form symbolic distillation method that efficiently converts the black-box GNN into a weight-free, analytically tractable symbolic equation. Compared to state-of-the-art approaches, our method achieves comparable or superior prediction accuracy while reducing parameter count by orders of magnitude. It thus enables lightweight deployment and delivers strong model interpretability—offering a novel paradigm for network performance optimization that jointly ensures high fidelity and full transparency.

Technology Category

Machine Learning: Graph-based Machine LearningPlanning, Routing, and Scheduling: Planning with Language ModelsData Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSocial Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Accurate prediction of flow delay is essential for optimizing and managing modern communication networks. We investigate three levels of modeling for this task. First, we implement a heterogeneous GNN with attention-based message passing, establishing a strong neural baseline. Second, we propose FlowKANet in which Kolmogorov-Arnold Networks replace standard MLP layers, reducing trainable parameters while maintaining competitive predictive performance. FlowKANet integrates KAMP-Attn (Kolmogorov-Arnold Message Passing with Attention), embedding KAN operators directly into message-passing and attention computation. Finally, we distill the model into symbolic surrogate models using block-wise regression, producing closed-form equations that eliminate trainable weights while preserving graph-structured dependencies. The results show that KAN layers provide a favorable trade-off between efficiency and accuracy and that symbolic surrogates emphasize the potential for lightweight deployment and enhanced transparency.

Problem

• Predict flow delay in communication networks
• Reduce trainable parameters while maintaining performance
• Create symbolic surrogates for lightweight deployment

Innovation

• Heterogeneous GNN with attention-based message passing
• FlowKANet integrates KAN operators into message-passing
• Symbolic surrogate models via block-wise regression distillation