| In recent years,with the development of communication technology,computing technology and hardware facilities,the Internet of Things technology has gradually penetrated into various fields of human daily life.In order to meet the needs of the global Internet of Things,it is necessary to establish a global communication network.However,due to the constraints of the geographical environment,it is often not economical to deploy ground base stations in remote areas.The LEO satellite communication system,due to its wide coverage and ability to break through the limitations of the geographical environment,can establish a satellite-based Internet of Things system as a supplement to the terrestrial Internet of Things.However,due to the heterogeneity between satellite network and terrestrial wireless communication network,it brings challenges to the design of satellite-terrestrial network fusion.Secondly,the random access of a large number of terminals and the different service quality of terminals of different service types bring great challenges to the access control and resource allocation of the LEO satellite Io T system,and even the terminal access requirements cannot be met.In order to solve the above problems,this paper focuses on three aspects: the satellite-terrestrial fusion network architecture based on Software Defined Network(SDN),the SDN controller deployment method in the LEO satellite network,and the SDN-based terminal access control and resource allocation method.The specific research contents are as follows:First,in view of the deep integration of heterogeneous networks in the coexistence scenario of satellite-terrestrial networks,this paper integrates the idea of SDN control plane and data plane separation,researches and designs the LEO satellite Io T architecture,and uses the SDN framework for access control and resource allocation.In addition,by adopting a flat distributed multi-controller architecture,the satellite node as a switch only needs to complete data forwarding and function configuration,while the satellite node where the controller is deployed focuses on access control and resource allocation decisions.To this end,the flexibility and scalability of the satellite network are improved,and the complexity of network configuration is effectively reduced.Secondly,aiming at the deployment problem of SDN controller of satellite network in the proposed architecture,a dynamic deployment algorithm of SDN controller based on Genetic Algorithm(GA)is proposed.Firstly,the problem of controller deployment in LEO satellite network is analyzed.Considering the time-varying network topology,a "snapshot" model of LEO satellite network is proposed,and a flat control architecture is adopted to provide faster communication performance between controllers..Then,the network delay is analyzed,and the optimization problem of controller deployment is established to minimize the network delay as the optimization goal.Finally,the genetic algorithm is used to solve the problem,and the possible deployment schemes are encoded in binary coding mode,and then crossover,mutation and other operations are performed to obtain the optimal deployment scheme iteratively.The algorithm proposed in this paper can realize the dynamic update of the controller deployment scheme with the service changes,and effectively reduce the satellite network delay.Finally,for the problem of preamble collision caused by terminal random access,an access control strategy based on dynamic Access Class Barring(ACB)mechanism is proposed.By dynamically adjusting the size of the ACB parameter and controlling the number of terminals that initiate random access,the purpose of reducing the collision rate and improving the access performance is achieved.On this basis,a multi-priority dynamic channel allocation algorithm is proposed to optimize the access performance of Io T terminals with different priorities.By dividing the limited channel resources and dynamically adjusting the channel resources allocated to each priority terminal,the access success rate of the terminal is improved. |