Research On Base Station Caching Strategy In Heterogeneous Networks

With the rapid development of mobile communication and mobile Internet,the user's requirements for mobile communication network have promoted the high-speed development of the fifth generation mobile communication system(5g).High data rate and low delay become the most urgent demand for mobile network.However,although traditional cellular network architecture adopts relay,OFDM,MIMO,antiinterference measures to increase system capacity and improve service quality,it still can not meet the growing needs of users and networks.The improvement and optimization of traditional network architecture has been paid more and more attention.The base station cache network emerges as the times require.Base station cache network has great advantages in operation cost,network coverage,network transmission rate and so on.In the real base station cache network,on the one hand,due to the limited cache capacity,how much cache capacity should be and what content should be cached has become the research difficulty and focus.On the other hand,in the direction of green communication,the base station cache network brings new storage cost and energy consumption.The time of cache update,the content of cache update,and the balance between energy consumption and system performance in cache network have become another focus of research.Firstly,the development of wireless communication system is introduced.The research status and main technical difficulties of the base station cache network structure are analyzed briefly,which lays the foundation for the subsequent research.Then,the technical foundation of heterogeneous cellular network is introduced.Firstly,the characteristics of heterogeneous cellular network are introduced and analyzed,and the differences of different base stations,signal coverage characteristics and SINR distribution in heterogeneous networks are introduced.Then,some characteristics of base station cache are introduced.The feasibility of the network popularity is introduced,and then the core technology cache strategy is described.The difference between active cache and passive cache is introduced.The current situation of mainstream cache strategy is introduced and the advantages and disadvantages are analyzed.Then,the green communication in the base station cache network is introduced.Three new green communication technologies,namely energy acquisition technology,flow sensing service supply technology and wireless multicast technology,are described in detail.Then,the main problems of this paper are: prediction,strategy and energy consumption optimization.Then,the optimization algorithm of cache hit rate for active cache strategy in base station cache network is introduced.A new user demand prediction algorithm based on deep learning is proposed.The algorithm aims to solve the problem of low hit rate and waste of cache performance due to dynamic change of popularity in network.Many articles about caching assume an ideal situation,that is,the popularity of the network is known,but in practice,this assumption is not true.In order to solve this problem,this paper transforms the focus of the problem from popularity in the network to each user in the network.The behavior model is established for each user's historical behavior,and the user behavior is learned by neural network,and the prediction is given.The model will follow the user behavior logic and constantly learn to change the behavior of the user,which is closest to the user's real behavior.The simulation results show that the prediction algorithm can effectively predict the behavior of users and greatly improve the cache hit rate.Then,a new set of joint cache strategy is proposed for the new prediction algorithm.The key of this strategy is to cache users in the base station service area,and cache the files based on prediction of user behavior and high popularity in the network.Based on the location model of user base station,the delay of user obtaining requirement file is simulated as the optimization target from the user location and actual demand,and the new joint cache strategy is compared with the traditional cache strategy.The simulation results show that the joint cache strategy based on user behavior prediction can effectively reduce the delay of users' access to the required files and improve the service quality of the base station cache network.Finally,aiming at the green communication problem advocated by 5g,the relationship between energy loss,user delay and economic value is analyzed under the combined cache strategy.Through the simulation of capacity and number of users,the results show that the conflict between providing the ultimate user experience and energy consumption,the lower the time delay of users obtaining files,the higher the energy consumption of the system.But for economic value,the system can choose to reduce QoS to ensure the maximum revenue.