Joint Graph-Based User Scheduling with LFoV Beamforming Design for B5G/6G NTN Systems

In this paper, we address the user scheduling problem for Low Earth Orbit (LEO) Multi-User (MU) Multiple-Input-Multiple-Output (MIMO) Non-Terrestrial Network (NTN) systems, aiming at maximizing the average per-user throughput while minimizing the total number of clusters. We introduce a graph-based scheduling approach utilizing Iterative Maximum Independent Set Search and Removal (IMISSR) with a constant graph density, combined with Limited Field of View (LFoV) Subarray architecture based on Signal-to-Leakage-andNoise Ratio (SLNR) beamforming. Users are clustered based on the channel coefficient of correlation (CoC) as a dissimilarity metric and served via Space Division Multiple Access (SDMA) using the proposed digital beamforming design on a cluster basis. These clusters are then allocated to different time slots using Time Division Multiple Access (TDMA). The results, presented in terms of per-cluster sum-rate capacity and peruser throughput, show that the combined IMISSR and LFoV beamforming approach greatly improves the system performance compared to conventional schedulers.