Research On Handover Technology In Satellite-ground Integrated Networks

With the rapid development and application of the Fifth-Generation Mobile communication technology(5G),relying solely on terrestrial networks can no longer meet the communication needs of end users for global wide area coverage in various scenarios.Therefore,there is an urgent need to study Satellite-Ground Integrated Networks supplemented by low Earth orbit satellite communication networks.Low Earth orbit satellite networks play an important role in Satellite-Ground Integrated Networks because of their lower latency,wider coverage and cost advantages compared with other satellite networks.However,its fastmoving nature leads to highly dynamic changes in network properties,and therefore requires corresponding user handover technology research for different scenarios.In this thesis,the corresponding handover process design and handover judgment algorithms are proposed to address the problems:timevarying property of the satellite-ground links,inefficient multi-user handover decisions,and the differences in the characteristics of the satellite-ground heterogeneous network.First,a Multi-Attribute Sensing Handover scheme(MSH)is proposed for the scenario of handover from a single user in a LEO satellite network.The scheme designs a handover signaling interaction flow for user-side sensing and adjudication,establishes a mathematical model composed of three handover factors,namely,reference signal received power,remaining service duration and remaining bandwidth resources,and performs weighting analysis by entropy weighting method.In this scheme,the multiattribute handover adjudication problem is transformed into a reinforcement learning problem,a reward function is constructed based on the handover factors and their weights,and the Q-learning algorithm is used for training and problem solving.Simulation results show that the scheme proposed in this thesis reduces the handover that exceeds the load threshold while guaranteeing the handover success rate,the handover rate that exceeds the load threshold of proposed scheme is 69%and 85%lower than that of the remain time based handover algorithm(Remain Time)and the reference signal received power based handover algorithm(RSRP),respectively.And the comprehensive evaluation quality of the proposed scheme respectively is improved by 11%and 3%compared with the Remain Time algorithm and RSRP algorithm.Secondly,for the scenario of simultaneous multi-user handover in LEO satellite networks,a Network Assisted Group Handover strategy(NGH)is proposed.The strategy performs user grouping and handover decision at the ground controller side,and designs the handover signaling interaction flow.A Constraint-based User Grouping Algorithm(CUGA)is proposed to group users according to their geographical locations and the bandwidth limitations of candidate satellites.In addition,a Maximum Flow based Handover Decision Algorithm(MFHDA)is proposed to make handover decisions for users within each group with the goal of maximizing the utilization of satellite bandwidth resources.Simulation results show that the strategy proposed in this thesis can reduce the number of handover times and handover blocking,as well as save signaling overhead in a user-dense scenario.Under the same conditions,the handover success rate is improved by 19%compared to the Remain Time algorithm and by 1.5 times compared to the RSRP algorithm.Finally,a Load Control based Satellite-Ground Handover strategy(LCSGH)is proposed for the scenario of simultaneous coverage of low Earth orbit satellite networks and terrestrial cellular networks.The strategy designs the signaling interaction flow of the satellite-ground cross-domain handover.A comprehensive utility adjudication function is designed for the user utility representations of different services and the overall network utility representations.It is followed that the ground network is the main one and the satellite network is the supplementary one,and a handover judgment flow is designed based on the load control threshold.Simulation results show that the proposed strategy can reduce the overall handover failure rate of the system by 14%and 13%relative to the simple additive weighting algorithm(SAW)and the RSRP algorithm,respectively,and at the same time improve the average throughput rate of the system.The average standard deviation of the load of the proposed strategy is reduced by 16%and 24%compared to the SAW algorithm and the RSRP algorithm respectively,which indicates that the proposed strategy can balance the load of the network to a certain extent.