Research On Cache Strategy In Cellular Heterogeneous Network

With the rapid development of wireless communication technology, the demands of users for data and multimedia services through the wireless network have a sharp increase. If we still adopt the traditional network optimization and management methods, it has been far unable to meet the explosive growth needs of current wireless traffic. At present, new communication technologies represented by MIMO, smart antenna and carrier aggregation technology have been widely developed and applied. Cellular heterogeneous network has many advantages, for instance, to enhance network coverage, to improve the transmission rate and the spectrum efficiency, flexible network and low cost etc, thus it becomes the inevitable choice of the future communications. But the cellular heterogeneous network still faces many technical challenges. On the one hand, heterogeneous cellular networks are still difficult to meet the huge demand of users, on the other hand, it also face more severe problem about backhaul. Small base stations(SBS) are usually connected with macro base station (MBS) through backhaul link. But the deployment of the backhaul link is expensive and the capacity is strictly limited, it will bring serious backhaul block when the demands of users increase. A promising technique for solving these problems is caching.Small base stations and the user equipments can download data and cache these data to direct service users. In this way, it can offer robust user's QoS,reduce burden of backhaul link and improve system capacity.This thesis focuses on the researches of cache strategies in cellular heterogeneous network, including two aspects as fellow:Firstly, in heterogeneous cellular network, small base stations download and cache the popular files when the network is idle. Small base stations are clustered based on popularity files to send demand files for users together by using the coordinated multiple points (CoMP) technology. The object is to maximize the system utility by designing reasonable beamforming and cache placement. The optimization problem is divided into two subproblems by exploiting the difference of optimization variables on the time scale, and it can be solved by proposed algorithm which joints beamforming and cache placement.Secondly, thesis considers cache strategies for mobile users in cellular network. The popular files can be cached into storage space of mobile user equipments. Users can obtain demand files through the local cache or D2D links from other mobile users, which can greatly reduce the transmission delay and improve user experience. The mobility of users is used to design and analysis cache strategies and the movement characteristics of any two mobile users can be described as Poisson process. By defining user satisfaction metric, the optimization goal is to maximize the average satisfaction of mobile user by mobility-aware user cache strategies. This constrained optimization problem is solved by a multiplier penalty function method. Then through the simulation results, the relationship between the user file demand probability distribution and the file cache probability distribution can be analyzed.