Research On Beam Hopping Mechanism And Resource Allocation For LEO Constellation

In the scenario of air-space-ground integration,satellite communication can be an effective supplement to the ground network.Due to the limited satellite resources,traditional single-beam communication can no longer meet the rapidly growing user needs.As a high-throughput satellite resource allocation method,beam-hopping technology can flexibly configure on-board resources according to ground service requirements,thereby improving on-board resource utilization.On the other hand,LEO satellite communication has great development potential in the future communication market due to its advantages of low construction cost,short propagation delay,and high communication rate.Due to the low orbit altitude and fast satellite movement,the distribution of user terminals and business requirements of low-orbit satellites are constantly changing.How to apply the beam-hopping technology to low-orbit constellation scenarios,design a reasonable and resource allocation algorithm,and achieve efficient matching between on-board resources and business needs has become an urgent problem to be solved.In this thesis,the beam-hopping mechanism of LEO is first studied.This thesis introduces the beam-hopping workflow of LEO constellations,constructs a typical LEO constellation scenario,introduces its forward link model,ground wave position distribution,and models user traffic.Based on the traditional access strategy,this thesis proposes a weighted access strategy.The simulation results show that the weighted access strategy can improve the average traffic angle without causing too much switching overhead.Secondly,under the condition that the network control center is the subject of resource allocation,this thesis establishes an optimization problem with the goal of maximizing the system throughput,and uses an iterative algorithm to solve the resource allocation problem.Since each beam reuses the on-board bandwidth resources at full frequency,this thesis uses spatial isolation technology to suppress co-channel interference between beams.Considering the uneven distribution of business demands,this thesis also optimizes the power allocation among beams.The simulation results show that compared with the traditional resource allocation method,the iterative algorithm proposed in this thesis can improve the system throughput and has better delay performance.Finally,due to the limitations of the iterative algorithm in the time dimension,this thesis proposes a resource allocation algorithm based on deep reinforcement learning in the case of satellites as the main body of resource allocation,and establishes an optimization problem with the goal of minimizing the queuing delay of data packets.The algorithm models the satellite as an agent,reads the state of each wave position data packet from the data packet buffer,and inputs the state into the DQN network after reconstruction,and then the agent decides to formulate a beam-hopping resource allocation plan.The simulation results show that,compared with the iterative algorithm proposed in this thesis and the traditional resource allocation method,the resource allocation algorithm based on deep reinforcement learning can reduce the queuing delay of data packets and improve the system throughput,and can adapt to the uneven distribution of traffic.