A Regularity-Driven Fast Rerouting Algorithm for Topology-Varying Satellite Networks

With the growing demand for global communication, Low Earth Orbit (LEO) satellite constellations have become essential for expanding network coverage and enhancing data transmission capabilities. Inter-satellite links (ISLs) are critical for enabling low-latency and high-throughput data transmission. However, frequent changes in satellite network topology due to ISLs failure or reconfiguration present challenges for routing efficiency. Traditional shortest distance path (SDP) routing algorithms are characterized by high computational overhead and slow convergence, limiting their applicability in dynamic satellite networks. In this paper, we propose a regularity-driven fast rerouting (RFR) algorithm that leverages the characteristics and regularities of ISLs in LEO constellations. To address ISLs failure or recovery, RFR adopts a local update strategy to quickly identify alternative shortest paths. Simulation results show that RFR outperforms the traditional Floyd-Warshall algorithm, substantially reduces routing update time, and can be effectively integrated with SDP-based routing algorithms.