Based On Distributed Small Base Station Cache Networks Content Placement Method And Service Strategy Research

In the era of 5G,base stations are densely deployed,causing backhaul congestion and time delay increase in the wireless networks.Edge cache can be used to solve the backhaul congestion problem caused by dense network and reduce network delay.Edge caching is caching popular files in base stations to avoid frequent information transfer from content providers to the edge of the network.When the network is densely deployed,few researchers will consider the situation that the user is in the state of high-speed movement,which is unreasonable in the context of dense network.In addition,in the edge cache of large heterogeneous cellular networks(HCN),it is also of high practical value to explore the optimal content placement strategy to provide better services for users.In mobile edge networks,edge cache supports the calculation and storage of data.Through the analysis and utilization of the collected user access data through the MEC server in the base station,the adopted caching scheme can be more suitable for the user's use environment.Therefore,the mobility-aware coded probability cache scheme is proposed for small base stations(SBS)supporting mobile edge computing(MEC),which integrates user mobility and distributed storage.This scheme combines user mobility,distributed storage and traditional probabilistic caching scheme to form a mobile-aware encoded caching scheme,and uses throughput to measure network performance.Based on random geometry theory and discrete random jump model,the throughput expression can be obtained.Because the expression is extremely complex,particle swarm optimization(PSO)and discrete particle swarm optimization(DPSO)are adopted to obtain the solution of the expression,and genetic algorithm is adopted to optimize the solution based on the solution for obtained the optimal solution.In addition,the effects of user mobility,content popularity and return coefficient on the revenue equilibrium of user mobility diversity,content diversity and channel selection diversity were determined by numerical analysis,and some basic conclusions obtained in SCN supporting MEC were applied to the recommended caching scheme.Finally,compared with the most popular cache(MPC)scheme and the conventional probabilistic cache scheme,the superiority of the proposed scheme is demonstrated.Numerical results show that,compared with the two schemes mentioned above,the proposed caching scheme can achieve higher throughput when the user is in a strong mobile environment and the file popularity is not skewed,and when the data link is requested,when the return link is poor in the SCN supporting MEC.Thisindicates that the proposed caching scheme can better solve the problem of network densification.Caching popular files in BS of HCN can avoid frequent information transfer from content provider to client,and can be used to reduce network delay and solve network congestion problem in backhaul link.Because the network performance analysis of optimal content placement strategy in large HCN is very complex,most existing strategies are based on approximation,heuristic or intuition,so the optimal content placement strategy in HCN still needs to be further studied.Using the most popular random HCN model,the k-layer model of BS is modeled as an independent poisson point process distributed on different density planes.Considering the random caching scheme,each file is placed on a certain layer of BS with corresponding probability,and the probability in BS of the same layer is the same,but the probability in different BS of the same layer is different.The hit probability is used as a measure of network performance,and the best layered placement strategy can be obtained by referring to the existing results of HCN network performance and maximizing the hit probability.When uniform receiving signal-to-interference signal(SIR)threshold values are adopted for different layers in BS to successfully transmit files,the optimal solution can be obtained.The placement probability of a specific file is proportional to the square root of the corresponding file popularity measurement,and its offset is determined by the BS cache capacity.For the general case of non-uniform SIR threshold,the optimization problem is non-convex.Therefore,the sub-optimal placement strategy is adopted by the approximation method.That is to say,the sub-optimal layout strategy is designed by the approximation method.