Research On Caching Strategy Based On Mobility And Content Request Prediction In Internet Of Vehicles

For intelligent transportation system,since the commercialization of the fifth generation cellular network,the number of content requests of vehicles has increased sharply.A large amount of data generated has a significant impact on the timeliness and stability of vehicle Internet(IOV)system,and vehicles have certain computing and storage capacity.Therefore,edge cache technology is applied to vehicle Internet.However,due to the limited computing and storage capacity of edge devices,it is difficult to ensure that all contents are cached,and all requirements of the device can meet all users.To solve this problem,this paper considers the content of vehicle request and vehicle mobility,and proposes an edge cache strategy based on mobility and content request prediction in the Internet of vehicles from the perspective of maximizing the hit rate of vehicle cache.The main research work of this paper is as follows:(1)Aiming at the problem that the limited caching and computing power of vehicles can not guarantee that all contents can be cached,an edge caching strategy of Internet of vehicles based on vehicle clustering and content request prediction is proposed.Firstly,the vehicles are clustered according to the speed and position of the vehicles to reduce the impact of vehicle mobility on the communication link and ensure the stability of subsequent content transmission;Then,based on the historical content request information,the number of requests for each content in the next time period is predicted through the LSTM network,and then the popularity of the content is obtained by using the Zipf model;Finally,reinforcement learning is used to solve the objective function to maximize the cache hit rate,and the optimal cache decision is obtained.Simulation results show that the proposed algorithm has higher cache hit rate and smaller delay than other algorithms,which shows the superiority and advanced nature of the algorithm.(2)Aiming at the problems of low cache hit rate and high delay caused by vehicle mobility.An edge caching strategy for Internet of vehicles based on vehicle clustering and mobility prediction is proposed.Firstly,vehicles with similar movement patterns are grouped by movement prediction,and then intra cluster and inter cluster communication is established by caching data in the selected cluster head.The vehicle periodically broadcasts beacon packets to one hop neighbors,and the beacon packets carry the movement information and status information of the sender.According to the clustering algorithm,vehicles can predict the future mobile mode through LSTM and link expiration time prediction model,and then establish vehicle clusters according to the prediction results.Finally,through cooperative caching,RSU selects some vehicles in the cluster head set as cache nodes to avoid the interruption of caching process caused by vehicle mobility.Simulation results show that the proposed algorithm has higher cache hit rate and less delay than other algorithms,which proves the feasibility and effectiveness of the cache scheme.