Research And Simulation Of Network Slice Orchestration Technology In 5G

5G will have higher user experience rate,greater connection capacity,extremely low end-to-end delay and other key applications,as well as a variety of application scenarios.5G will also comprehensively improve the openness and innovation of services,and significantly reduce the cost of network construction and operation.In order to meet the 5G service requirements,5G networks introduce technologies such as NFV and network slicing to drive network architecture innovation.The network slicing technology is very important in 5G because of its flexible configuration,resource sharing,scalability and support for vertical industry applications.How to design and orchestrate network slices to achieve optimal service quality and performance has become the core and difficulty of current 5G network slice research.In this thesis,the research on network slicing and scheduling technology in 5G environment is mainly carried out from two aspects.First,how to arrange and manage the access network slice resources on the access network side,while ensuring the isolation of the network slice and improving the utilization of the spectrum resources in the access network;second,how to be on the core network side,The resources required for slicing in the core network are reasonably arranged and allocated,so that the core network slice has the best network performance,and the infrastructure resources can carry more network slices to improve the network operator's benefits.In response to the above problems,this topic studies the progress and challenges by analyzing the above problems based on the development characteristics of 5G and future mobile communication networks.The final decision was to focus on the study of network slicing in the 5G core network.In the course of the research,the following work was mainly completed:First,a study on the layout of 5G network slices was conducted.A new network slice mapping model is designed based on the optimization goal of the network slicing scheme in the existing research and the difference in the performance requirements of the slice service network.The slice operation cost and the slice are introduced in the model.The service type changes the network performance value,so that the generated network slice can reduce the operating cost of the network operator and enable the network slice to have better network performance.Furthermore,according to the characteristics of the model,a mapping algorithm based on genetic tabu search algorithm is designed to solve.The algorithm can search the mapping position of the global optimal VNF and virtual link in the underlying network in parallel according to the historical memory.Secondly,based on the network slicing scheme designed in this thesis,combined with 5G network slicing and NFV MANO architecture,a network slicing programming simulation system is designed,which is a concrete realization of the 5G network slicing function.Finally,the above network slicing programming simulation system is realized,and the feasibility of the network slicing scheme and system designed in this thesis is verified by simulation experiment and functional test.