Research On Satellite Internet Routing Algorithm

Satellite Internet that has emerged in recent years refers to a system that provides Internet access services to the world based on communication satellite constellations.It has the advantages of wide coverage,low propagation delay,and high communication speed.Providing high-quality network support in areas where the terrestrial network is difficult to cover,effectively eliminating the blind spots of network coverage,Satellite Internet is an important strategic infrastructure in the future.In the satellite Internet related technologies,the routing algorithm has an important impact on the overall performance of the system.Therefore,the Satellite Internet routing algorithm has important research significance.Compared with traditional communication constellations,the routing algorithm of Satellite Internet constellations faces the following challenges:Satellites are susceptible to failures due to some factors such as space weather,due to the huge size of the constellation,the probability of Satellite Internet failure is higher,therefore,there are higher requirements for the robustness of the routing algorithm;Satellite Internet faces the challenge of integrating with the terrestrial network to realize the integration of the satellite-terrestrial network,and the existing satellite network routing method is difficult to integrate with the terrestrial network;The data traffic carried by Satellite Internet is huge and unevenly distributed,which may cause link congestion easily and reduce quality of service.In order to solve the above problems,this thesis conducts research on Satellite Internet routing algorithms.The main research contents are as follows.(1)Optimized Routing Request Area based Location-Assisted Ondemand Routing algorithmAiming at the meeting high requirement of robustness of Satellite Internet routing algorithms,an Optimized Routing Request Area based Location-Assisted On-demand Routing algorithm(ORRA-LAOR)is proposed.This algorithm updates the routing table through routing request and routing reply messages,and adopts a newly designed routing request area,which makes targeted optimizations to the original routing request area for the situation that LAOR is prone to the problem of reduced robustness.While improving the robustness of the algorithm,the number of routing request messages is reduced.The simulation results show that in the case of any single satellite failure,ORRA-LAOR can successfully discover the route.In the case of any number of satellite failures,the route discovery failure rate of ORRA-LAOR algorithm is lower than LAOR,and,the average routing cost of ORRA-LAOR is about 33.86%lower than that of LAOR.(2)Load-balanced routing algorithm based on geolocation coded IPIn order to solve the problem of the difficulty in integrating the routing methods of satellite-terrestrial network,a Load Balancing routing algorithm based on Geolocation Coded IP(LBGC)was proposed.This algorithm uses the earth’s surface block coding algorithm to convert the coordinates of the terminal geolocation into binary codes and embed them in the IP address.Each satellite can calculate the destination terminal position through the destination IP address,and comprehensively calculate and select the forwarding interface and update routing tables according to the relative position and interface cache level.The results of theoretical analysis and simulation experiments show that the algorithm consumes very little computing and storage resources.Compared with the discretetime dynamic virtual topology routing(DT-DVTR)algorithm,the standard deviation of its load distribution decreased by 42.46%.(3)Traffic Prediction based Satellite Internet Load Balancing algorithmIn order to solve the problem that the satellite Internet carries a large amount of traffic and is prone to link congestion,a Traffic Prediction based Satellite Internet Load Balancing algorithm(TPLB)is proposed.The algorithm first collects historical traffic data,divides the historical data into high-load periods and low-load periods according to the periodicity of satellite traffic,and uses Long Short-Term Memory(LSTM)network to predict traffic in different periods.Then,using the predicted traffic data,according to different routing algorithms,a centralized or distributed traffic adjustment strategy is used to adjust the routing results to avoid congestion as much as possible.The simulation results show that the LSTM segmented traffic prediction method improves the accuracy by 49.6%compared with the original LSTM network.Through traffic adjustment,TPLB effectively improves the load balancing performance of the network and reduces the network congestion rate.