Handover Algorithm Based on User Clustering and Graph Matching for Large-Scale LEO Satellite Network

This paper proposes a Customized Satellite Handover Algorithm based on User Clustering and Graph Matching (C-UCGM), aimed at addressing the challenge of simultaneously ensuring user QoS preference and load balancing in largescale LEO satellite network. The algorithm identifies LEO satellite access points and controllers through satellite clustering, achieves customizable user services through multi-criteria weighted decision-making, and employs an algorithm based on user clustering and graph matching to determine the target LEO satellite for handover. Protocol procedures are utilized to ensure the handover process. These methods enable us to meet user QoS requirements while determining handover relationships according to LEO satellite loads. Simulation results demonstrate that, compared to traditional handover strategies, C-UCGM can reduce handover frequency by 12.6 %, increase the handover success rate by 21.9 %, and load more balance among LEO satellite. This signifies a notable improvement in both operational efficiency and service quality within large-scale LEO satellite networks.