Service-Oriented Network Slicing For Vehicular Communication Network

With the explosive growth of mobile traffic,mobile network operators are facing severe challenges in providing the required capacity growth.In addition,the 5G and future networks will mainly face vertical industry users and support multiple use cases,so the services it needs are extremely diverse.Therefore,mobile network operators have proposed network slicing technology to enable multiple virtual networks to share the underlying infrastructure and wireless resources to support flexible and diverse scenarios and services.However,the current network architecture does not take into account the space-time differences between application scenarios,resulting in extremely uneven use of wireless resources,and network devices will become idle.Therefore,there is an urgent need to design new network architectures and management methods to achieve efficient and differentiated use of wireless network resources.First,the research background of network slicing and vehicular communication network and application scenarios of vehicle communication network are introduced.The research purpose and significance of vehicle network slicing are briefly introduced.Then,it analyzes the research status of vehicle network slicing and virtualization and its key problems.Secondly,a slicing coordination agent based on service guarantee is proposed.Use the K-means ++ clustering algorithm in the service clustering module to cluster V2 X communication services and map them to different slices.In the slice scheduling module,use the shared proportional fairness scheme to improve wireless resource utilization and consider service requirements,design a resource allocation algorithm based on linear programming obstacles to get the optimal slice weight allocation result.Thirdly,a slicing resource allocation strategy based on machine learning is proposed.Conv LSTM captures the temporal and spatial correlations of different business traffic to predict business traffic.The prediction result is substituted into the modeled total system delay problem,and a resource allocation algorithm based on the original dual interior point method is designed to obtain the optimal slice weight allocation result.Finally,the research content of the full text is summarized,and the subsequent research square is prospected.