LEO satellites move at high speeds relative to the ground,resulting in frequent switching of ground terminals between different beams of the satellite or between different satellites,affecting peopleās communication experience.The service-oriented architecture of 5G communication core network and mature handover management can provide a reference for LEO satellite communication handover.Therefore,this paper optimizes the ground terminal handover process based on the 5G satellite base station network architecture,analyzes relevant parameters and motion models of LEO satellite,and aims at the poor comprehensive performance of the current LEO satellite handover algorithm due to the lack of adaptive adjustment according to Satellite Network changes,The main tasks of this article include:Optimize the handover process of terminals in the satellite ground fusion network based on the 5G satellite base station architecture,so that LEO satellites can submit handover processing requests to the ground core network for processing.For intra satellite beam switching,a dynamic channel reservation algorithm is used to balance the new call requests and switch requests of the terminal,and the number of reserved beam channels is dynamically adjusted based on changes in the terminal switch request rate and new call request rate;For inter satellite handoff,the fuzzy analytic hierarchy process is used to comprehensively consider the reference signal reception quality,inter satellite link channel signal-to-noise ratio,remaining time,and number of idle channels,to solve the optimal handoff access satellite,and to increase the adaptive adjustment process to dynamically adjust the decision attribute weight according to the changes in Satellite Network.Through simulation,it is verified that the handoff times,failure rate and other indicators have good performance.Further migrate the AMF function of the 5G core network to the onboard base station,making it the main control satellite for managing the switching management of the onboard base station,enabling switching requests to be handled by the onboard AMF without the need to transmit them back to the ground core network.In response to the problem of selecting the number of main control satellites during dynamic satellite based base station switching,the MADDPG algorithm is used to optimize the placement of AMF on the LEO satellite side.The number of main control satellites and the number of AMF instances are calculated based on the change in handover request arrival rate.The optimal main control satellite is selected by combining the adaptive multi-attribute decision-making inter satellite switching algorithm,and an experimental simulation environment is built to test performance indicators,verifying the feasibility of the scheme. |