Dynamic User Association in Scalable Ultra-Dense LEO Satellite Networks

Ultra-dense low earth orbit (LEO) satellite networks (LEO-UDSNs) are becoming a foundational component of future 5G and beyond. LEO-UDSNs bring unprecedented high-speed mobility, broadband capacity, and ultra-density qualities, introducing a significant challenge to traditional user association (UA) methods. In this article, we first discuss the difficulties in sequential UA optimization for scalable LEO-UDSNs. A specific multi-objective UA problem was formulated for the overall service efficiency. We initially proposed a practical Iterative Subgradient Algorithm (MOISA) decoupling potential trade-offs between limited equivalent power and availability relationship, rapidly obtaining solutions in stable cases. Second, we proposed a Multi-Objective Dynamic User Association Approach (MODUAA), inheriting the MOISA with stability enhancements for sequential optimizations, ensuring the performance requirements, and alleviating the iteration redundancy in dynamic scenarios. Finally, the outperforming of the MOISA and the applicability of the MODUAA were elaborated in various comparisons.