Routing Convergence Overhead Measurement And Optimization For LEO Satellite Networks

In recent years,with the rapid development of satellite technology and the proposal of the 5G Internet of Things concept,the development of satellite networks has attracted much attention.Low Earth orbit(LEO)satellite networks have a shorter operating cycle and more flexible coverage,making them an important research direction in satellite networks.For example,SpaceX’s Starlink program aims to build a largescale LEO satellite network to provide high-speed internet services to users.In the complex environment of outer space,the continuous expansion of LEO satellite networks has made the problem of link failure between satellites more prominent.How to quickly and efficiently route and transmit data packets has become a bottleneck issue for the expansion of LEO satellite networks.At the same time,the inter-satellite link bandwidth in LEO satellite networks and the computing and storage capabilities of satellite nodes are limited.How to combine LEO satellite networks with dynamic routing algorithms under limited resources is a hot issue in related research at home and abroad.This thesis aims to solve the above problems.First,based on the Linux system and combined with Docker container technology,a highly flexible satellite network routing simulation platform is built.When the satellite network link is disconnected,the Open Shortest Path First(OSPF)protocol algorithm is used for route convergence to update the routing table and measure route convergence overhead.The measurement results show that as the scale of satellites increases,the route convergence time shows a rapid growth trend,and the delay fluctuations caused by satellite network business flows also increase.In response to these phenomena,this thesis proposes a route convergence optimization scheme based on OSPF routing protocol.The route convergence optimization scheme takes advantage of the existence of redundant routes in satellite networks and defines the flooding area of topology information when links are disconnected in OSPF routing protocol within a certain range.Satellite nodes within the range flood normally,while those outside do not update topology information.Since boundary satellites can route correctly,it can ensure the correctness of routing throughout the satellite network while reducing flooding overhead and convergence time for route convergence.The local flooding scheme reduces overhead without ensuring that satellite network routes are shortest,resulting in an increase in hop count for some routes.After simulation experiments and measurements,it was found that in largescale satellite networks,the increase in hop count brought about by increased overhead can be ignored compared to reduced flooding overhead.The simulation results of the route convergence optimization scheme show that the optimization scheme can effectively reduce route convergence time and reduce overhead caused by route flooding.At 48star scale,convergence time was reduced by 35.11%and LSU packet count during flooding convergence was reduced by 96.06%;at 200-star scale,convergence time was reduced by 87.23%and LSU packet count during flooding convergence was reduced by 99.10%.At the same time when a link is disconnected in a satellite network,the maximum increase in hop count for routing paths is 2 hops.At the same time when a link is disconnected in a satellite network,the maximum increase in hop count for routing paths is 2 hops.According to these experimental results,it can be concluded that the proposed route convergence optimization scheme of this thesis has low flooding overhead and fast convergence time,while ensuring routing correctness in satellite networks.This thesis provides ideas for solving excessive flooding overhead caused by large-scale dynamic routing protocols,and has some significance for research on large-scale satellite network routing algorithms.