Research On Caching Technologies Of Wireless Services In High-speed Mobile Scenarios

With the rise of Internet,the popularity of intelligent terminal,the development of the broadband multimedia services and the high-speed train travel demand increase,people in high-speed mobile scenario has more and more communications needs.However,due to the doppler frequency shift and high-speed train special material body loss in the high-speed scenario.The available data bandwidth per passenger is far less than the ground motion at low speed or the stillness of the average user bandwidth.Therefore,the demand for wireless data bandwidth is more urgent in high-speed mobile scene.How to improve the data bandwidth of each passenger in high-speed mobile scene is more urgent.Therefore,this paper studies the wireless services caching technologies in the converged wireless broadcast and cellular network of high-speed scenario,focusing on the wireless caching strategy and algorithm when the high-speed railway has different sites.First of all,in the case of only two stations on a single railway line,we study how to maximize the equivalent throughput of the wireless converged broadcast and cellular network in the limited time.Thus,the wireless converged broadcast and cellular network model is proposed in this paper.And change the services cache problem of the two-stations into the0-1 knapsack problem.This paper adopt the dynamic programming algorithm,size-based greedy algorithm,equivalent throughput-based greedy algorithm,popularity-based greedy algorithm to solve the 0-1 knapsack problem.Simulation results show that the performance of the popularity based algorithm is close to the optimal solution-dynamic programming algorithm,size-based algorithm take the second place,equivalent throughput-based algorithm is worst.Next,the complexity of these four algorithms is analyzed quantitatively.Analysis shows that the time complexity of the proposed greedy algorithm is far less than dynamic programming,which is the optimal solution.The upper and lower bounds of the popularity greedy algorithm performance is proved in this paper.This paper also quantitatively analyzes the performance of each algorithm in the case of different Zipf factor and different minimum services size.And analysis shows that: popularity based algorithm is consistent with the simulation performance.Then,the repeated push strategy of services in the converged wireless broadcast and cellular network was studied in the case of three stations or multiple stations when passengers getting on and off.Based on the characteristics of the long tail distribution,the number of passengers on the train is modeled and the u-shaped distribution is used to describe the number of passengers on different stations getting on and off.We firstly refine the services collection to be re-pushed,as the high-speed train arriving at the intermediate station,we may have some passengers getting on and off,then according to the number of people getting on and off is required to adjust the wireless services strategy to re-push the cached services.By comparing the equivalent throughput of the rebroadcast strategy and not rebroadcast in the converged network,to decide whether to re-push the hot services.The simulation shows that the proposed re-push strategy can effectively improve the equivalent throughput of the system,especially in the case of passengers on the train all get off the train in the intermediate station,the improvement of the equivalent throughput can reach 145%.Then,in view of the short time of high-speed train stopping at the intermediate station and the limited reception capability of vehicle relay station,this paper studies the problems of how to make the vehicle relay station get more new hot services during the limited parking time.In this paper,the crowd sensing networks is introduced,and the reverse auction model is introduced too.In order to stimulate the enthusiasm of the terminal users in the carriage.The concept of social welfare is introduced to measure the effectiveness of the algorithm.The objective function of maximizing social welfare is established.In order to find an approximate solutions in polynomial time,basing on the value characteristics of hot services,an auction algorithm based on popularity is proposed.Simulation results show that the proposed popularity auction algorithm outperforms the existing algorithm in terms of social welfare and task completion ratio.Finally,this paper summarizes the above research work and puts forward some future research directions.