Toward Scalable SDN for LEO Mega-Constellations: A Graph Learning Approach

2026-04-30

AI Summary

This study addresses the scalability challenges in managing large-scale low Earth orbit (LEO) satellite constellations, where conventional software-defined networking (SDN) struggles to efficiently control thousands of satellites interconnected via inter-satellite links. To overcome this, the authors propose a hierarchical SDN framework that uniquely integrates graph neural networks with Koopman operator theory, introducing a Graph Koopman Autoencoder (GKAE). This architecture enables topology compression and linearization of nonlinear dynamics within orbital shells, facilitating efficient spatiotemporal behavior prediction. Global coordination is achieved through aggregation by a central controller. Evaluated on a Starlink-like constellation, the approach achieves over 42.8% improvement in spatial compression ratio, a 10.81% gain in temporal prediction accuracy, and a significantly reduced model size compared to baseline methods.

Abstract

Terrestrial network limitations drive the integration of non-terrestrial networks (NTNs), notably mega-constellations comprising thousands of low Earth orbit (LEO) satellites. While these satellites act as interconnected network switches via inter-satellite links (ISLs), their massive scale creates severe bottlenecks for network management. To address this, we propose a scalable, hierarchical software-defined networking (SDN) framework. Our architecture leverages graph neural networks (GNNs) to compactly represent the constellation topology, and Koopman theory to linearize nonlinear dynamics. Specifically, a Graph Koopman Autoencoder (GKAE) forecasts spatio-temporal behavior within a linear subspace for each orbital shell. A central SDN controller then aggregates these shell-level predictions for globally coordinated control. Simulations on the Starlink constellation demonstrate that our approach achieves at least a 42.8\% improvement in spatial compression and a 10.81\% improvement in temporal forecasting compared to established baselines, all while utilizing a significantly smaller model footprint.

Problem

• LEO mega-constellations
• scalable SDN
• network management bottleneck
• inter-satellite links
• constellation topology

Innovation

• Graph Neural Networks
• Koopman Theory
• Software-Defined Networking
• LEO Mega-Constellations
• Graph Koopman Autoencoder