Demand- and Topology-Aware Resource Allocation in Non-Terrestrial Networks (NTNs) with Multi-Satellite Beam Hopping

NGSO constellations, represented by LEO and MEO satellites, require resource-allocation frameworks that jointly address payload flexibility and communication performance. This work presents a hierarchical three-stage framework for resource allocation: demand-driven satellite cell coloring (SCC), topology-aware satellite–cell association (S2C), and multi-satellite beam hopping (BH). Our framework halves the downlink power required to reach 10% unserved capacity (UC) and reduces the peak load on inter-satellite links (ISLs) by approximately 40% compared to a baseline strategy, while controlling cell handovers. The results further show that architectural parameters dominate performance: increasing the number of beams, refining cell granularity, and scaling constellation density enable approximately 2×, 10×, and 50× reductions in power consumption, respectively. These reductions in the required downlink (DL) power and inter-satellite link (ISL) load for a given performance target can be directly translated into system-level gains at payload or constellation size level. The proposed framework provides a scalable foundation for end-to-end optimization of next-generation NGSO satellite networks.