Research On Caching Technology In Heterogeneous Cellular Networks

To support the burgeoning demand for mobile data traffic,the next-generation cellular network will appear as a heterogeneous multi-cell cellular network consisting of macrocell base stations and dense small cell base stations.In heterogeneous cellular networks,the stringent backhaul capacity of small cell restricts the further improvement of network performance.Caching as an effective solution has become the current research hotspot.By deploying cache in small cell base stations and pushing popular files to cache in advance,users can obtain files directly from the cache without passing the backhaul,thereby effectively improving the network performance.In this paper,the small cell caching strategy is optimized for different scenarios and optimization objectives.Firstly,the optimization of the small cell caching strategy with the goal of maximizing the hit probability is studied.Previous researches have discussed the optimization of the small cell caching strategy based on the hit probability in the scenario with small cell idling and using the information-centric association(ICA).On that basis,this paper analyzes the caching strategy optimization problem in the scenario without idling of small cell and using ICA.The problem is simplified based on the assumption that the user density and small cell density are much greater than macrocell density,and then a "two-step optimization method" is designed to transform the original problem into two convex optimization problems.In addition,this paper further studies the caching strategy optimization problem for using the connection-based association(CBA).It is proved that the most popular caching(MPC)is the optimal caching strategy.Simulation results show that the optimal hit probability of using ICA is greater than that of using CBA.Secondly,the optimization of the small cell caching strategy with the goal of maximizing the area spectrum efficiency is studied.In the previous researches,the optimization of small cell caching strategy based on the area spectrum efficiency has been discussed in the scenario with small cell idling and using ICA.Based on that,this paper analyzes the optimization of the caching strategy in the scenario without idling of small cell and using ICA.It is proved that MPC is the optimal caching strategy based on the assumption that the cache capacity is large enough and the user density and small cell density are much greater than the macrocell density,as well as the approximation of the Gaussian hypergeometric function.Furthermore,this paper studies the optimization problem of the caching strategy for using CBA.It is also proved that MPC is the optimal cache strategy based on the approximation of Gaussian hypergeometric function.Simulation results show that using ICA does not necessarily have greater area spectral efficiency than using CBA.Finally,the optimization of the small cell caching strategy in the scenario with retransmissions is studied.At present,some scholars have studied the impact of retransmissions on the caching strategy of base stations in a single-tier network.This paper further studies the situation in a two-tier heterogeneous cellular network based on that.For the case of mobile users and CBA,it is proved that the optimization problem of the hit probability is a convex optimization problem.For the case of mobile users and ICA,a “two-step optimization method” is designed to convert the original problem into two convex optimization problems,based on the assumption that the user density is much larger than the small cell density while the small cell density is much larger than the macrocell density.For the case of static users and CBA,it is proved that MPC is the optimal caching strategy.It is discovered that for mobile users,the small cell base stations tend to cache the most popular files when the maximum number of transmissions is small,and tend to adopt a more uniform caching strategy when the given maximum number of transmissions is large.Besides,the simulation results show that user mobility can increase the diversity of the cached files and improve the hit probability.