Research On Access Slice Selection Technology In Heterogeneous Vehicular Networks

With the rapid development of mobile communication and Internet technology,as one of the three application scenarios of 5G,the Internet of Vehicles has become a new industrial form with deep integration of transportation and information industries.However,with the increasing demand for vehicle applications and vehicle data transmission,traditional single Internet of Vehicles access technologies cannot support their incremental emerging service types and data size,and heterogeneous vehicular network access technology remains its key technology.One of the k ey technologies of 5G networks includes Network slicing technology.It can provide the logical network functions required by specific application scenarios to support the transmission needs of different services.This paper introduces network slicing technology into the Internet of Vehicles,and achieves the goal of improving the QoS of the Internet of Vehicles.It is of great significance to improve the level of automation and intelligence of automobiles,realize automatic driving,and develop intelligent transportation.Firstly,based on the research of communication access technology in vehicular networks,this paper studies the Internet of Vehicle access and congestion problems on the WAVE and cellular networks system.This paper constructs a heterogeneous vehicular network access architecture based on network slicing,and conducts in-depth research on the selection of access methods and QoS ooptimization for heterogeneous vehicular networks.Research shows the heterogeneous network alleviate the network congestion problem and improve the QoE of the vehicle users effectively.Then,this paper analyzes the vehicular network access technologies such as DSRC technology,LTE-V2X technology and WiFi technology,and discusses the slice selection strategy of access technology for different types of application service requirements under different network slices.A one-to-many matching algorithm is proposed,and a reasonable one-to-many mathematical model is established.The simulation analysis is compared with the method based on multi-attribute decision and Bayesian prediction.The simulation results show that the one-to-many matching algorithm can effectively improve the system throughput under the condition of satisfying the user's QoE demand,and can reduce the number of handovers and reduce the handover delay.At last,in order to improve the resource utilization and network load balancing performance of the Internet of Vehicles,this paper studies the selection of different access slice in the heterogeneous vehicular networks based on the reinforcement learning algorithm,and proposes a Q-Learning access slice selection strategy for different QoE requirement,SNR conditions and vehicle speed.Through its intelligent data collection,data analysis and data processing,the appropriate access slice is obtained.The simulation results show that the proposed algorithm effectively improves the system resource utilization under the premise of balancing load balancing between different networks.Further research directions can be combined with channel conditions,vehicular networking business characteristics and driving speed to study the vehicle data access selection strategy and optimization algorithm based on big data to improve system QoS.