| Low Earth Orbit(LEO)satellite communication is a very important component in non-terrestial wireless communication networks.The broken down of Inter-Satellite Link(ISL)and the relationship between satellites bring out changes of the topology in a LEO satellite network,and this induces the update of snapshot in the conventional Virtual Topology(VT)technology.Existing VT schemes are difficult to address the issue of fast snapshot updates.When an access User Equipment(UE)handover among satellites,those control signaling transported over a loosy ISL suffers longer latency,and even time-out failure.This thesis focuses on interstellar routing and handover techniques.The main contribution of the thesis is as follows.Firstly,LEO satellite constellation,interstellar routing and handover in service link are overviewed,and rerouting of inter-satellite path and high-concurrency of handover are focused on.The typical topology of LEO satellite network is obtained by the existing VT scheme and topology sanpshot.A Waypoint-based Segment Routing(WSR)algorithm is proposed to avoid such loosy ISL of satellite in entering into Earth polar area by selecting a waypoint satellite and then routing packets segmentally.Then,in order to reduce handover latency and failure rate due to rerouting and high-concurrency,an optimization problem model with the goal of minimizing the average handover latency is established.Based on the model and WSR,the handover scheme is designed.Based on the existing channel models of ISL and Service Link,formulas for theoretical calculation are derived for the optimization model.Finally,to verify the applicability of WSR,a network simulation of the algorithm is designed and implemented over NS-3,and then comparative experiments are conducted with OLSR in two test cases.Compared with OLSR,WSR reduces network jitter by 46 ms and 126 ms,respectively.To evalutate performance for the proposed handover scheme,a numerical simulation is implemented by MATLAB.The simulation results show that handover latency reduce in average by 31.4%. |