Massive Beam Scheduling in LEO Systems: Low Complexity via Effective Interference Approximation

Multi-beam Low Earth Orbit (LEO) satellites are poised to play a central role in integrated 6G in space. Existing approaches for beam scheduling, with exact inter-beam interference modeling in every time slot, greatly suffer from prohibitive computational complexity as the number of beams scales up. This paper introduces a tractable interference approximation model that captures inter-beam interference in an aggregated form, completely removing slot-level computations while maintaining high accuracy. The model admits efficient numerical solutions that yield the probabilities of beam activation, enabling near-instantaneous construction of beam scheduling with minimum computational cost. Simulation results show that our method maintains over 95% accuracy in predicting scheduling performance compared to exact interference modeling, with less than 6% loss relative to the global optimum for capacity-demand ratio optimization. These results highlight the effectiveness of the proposed model as a foundation for efficient beam scheduling in next-generation large-scale LEO satellite systems.