Beam-hopping Pattern Design And Channel Allocation In LEO Satellite Communication System

LEO(Low Earth Orbit,LEO)satellites have a wider coverage area and fewer geographical restrictions compared to cellular networks,and have the advantages of low latency,low cost and low loss compared to high-orbiting satellites,which play an important role in communication scenarios such as ocean,mountainous areas and emergency situations,and are an important part of global mobile communication systems.LEO satellite communication system is a typical resource constrained system,due to the non-uniform distribution of ground users,the traditional fixed allocation scheme is difficult to effectively improve the user service quality.The beam-hopping technology enhances the resource utilization with the idea of time slicing,which pursues the goal of suppressing interference and enhancing capacity,but serious interference occurs when the beams of the same channel are close to each other.The design of beam-hopping pattern suppresses interference by optimizing the distance between beams in the same channel.The channel allocation dynamically schedules the channel resources according to the user service distribution to adapt to the service differences between beams.The research on beam-hopping pattern design and channel allocation is of great significance to improve the resource utilization and user service quality of LEO satellite communication system.This paper introduces a centralized resource pool architecture for LEO satellite communication systems,design beam-hopping patterns and dynamically assign channels,with the following main research elements:(1)A beam-hopping pattern optimization strategy based on Lagrange algorithm is proposed to take into account inter-beam interference and system capacity.Firstly,this strategy models the beam-hopping time slot allocation problem with the goal of inter-beam service fairness and solves the number of time slots for each beam;secondly,maximizes the number of working beams in the time slot by considering the interference distance when designing the beam-hopping pattern,and gives the specific design flow.The experimental results show that the strategy can ensure the fairness among the beams and satisfy the user requirements to the maximum extent(2)A channel allocation strategy based on reinforcement learning is proposed to solve cell edge user interference and adapt to inter-beam service variability.Firstly,this strategy forms a NOMA(No-Orthogonal Multiple Access,NOMA)cluster of edge users and central users to realize power multiplexing,and develops a pairing scheme between edge users and central users according to the location of edge users;secondly,abstracts satellite communication system,user service requests and channel resources as agent,state and action respectively,designs a reward function according to the channel capacity of users,and uses reinforcement learning algorithm to train a reasonable channel allocation strategy,furtherly,the update strategy is improved to accelerate the convergence of the algorithm.Simulation results show that the strategy reduces the outage probability of edge users and improves the performance in system blocking rate by 14% and 18% relative to Lagrange channel allocation and fixed channel allocation,respectively,which effectively improves the user service quality.