Research On Routing Strategy Of LEO Satellite Internet Of Things

In recent years,the Internet of Things has attracted great attention from many countries.Scholars from all over the world are exploring the broad application prospects of the Internet of Things.With the arrival of Internet of Things,many IoT application scenarios are gradually being implemented,including ocean monitoring,weather warning,and so on.At the same time,large-scale IoT terminal equipments have been put into remote areas such as deserts and oceans for real-time data transmission.As an effective means of communicating remote areas,LEO satellite networks will become an indispensable pillar in the sky-integrated system of the Internet of Things.LEO satellite networks have many advantages such as global coverage and short delay.If LEO satellite network and the Internet of Things are combined,the communication problem can be solved for the IoT terminal equipment.In recent years,many routing algorithms have emerged for LEO satellite networks,but most of them are not targeted at IoT scenarios,and their performance in congestion control is not ideal.Firstly,this paper studies the K shortest path algorithm in the Iridium model,and simulates the massiveness and suddenness of the IoT terminal.The result shows that the K shortest path algorithm is unable to balance the load,causing some link congestion to affect the overall performance.Then,K shortest path algorithm is improved.The original simple avoidance of the congestion route is changed to the weighted way.Through the simulation,the result shows that the packet loss rate has decreased by the improvement.After that,the process of rerouting is added.The simulation results show that the packet loss rate and throughput performance are improved after the rerouting process is added..However,under the large-scale constellation model,such as the OneWeb-like constellation,if the K shortest path routing algorithm based on the virtual topology is adopted,the routing table will be too redundant because too many nodes in the constellation,thereby reducing the computational efficiency.Therefore,distributed algorithms are a good choice in large-scale constellations.In the paper,the LCPR algorithm is studied in the OneWeb-like constellation,and the simulation is done for the IoT scenario.The results show that LCPR which only considers the delay within two hops,lacks the overall understanding,so it is easy to be congested in individual nodes.LCPIR algorithm is proposed by improving LCPR algorithm.The packet loss rate and throughput of the algorithm are simulated and analyzed under the class OneWeb constellation.The results show that both performances are improved.At last,a method of simulating satellite routing algorithm with STK+Matlab is proposed.The construction of experimental environment,experimental methods and the influence of different parameters on simulation results are introduced in detail.