A Research On Cache Technology In Multi-antenna Ultra-dense Network

With the rapid development of mobile communication technology,the data traffic of the network is growing exponentially,but the limited spectrum resources retard the improvement of the performance of traditional cellular network.In order to solve this problem,the ultra-dense network model of densely deploying small base stations in Macro stations has been widely studied and applied.Small base stations have the advantages of low power consumption,low cost and easy to deploy,which can make up for the coverage blind of Macro stations and effectively improve the system capacity.At the same time,because the multi-antenna technology can improve the utilization of spectrum resources effectively,increase the data transmitting rate exponentially,and improve the transmission quality greatly,it is valuable to adopt the multiantenna technology in the wireless network.However,due to the dense deployment of nodes and the large amount of data transmission in the ultra-dense network,the backhaul congestion is aggravated,which will lead to network delay and affect the quality of user experience.Therefore,how to reduce the link congestion becomes an urgent problem.As the key technology of5 G,cache technology stores popular files in the base station during off-peak period,when a user requests a file,user can directly obtain the required resources from the cache of base station,instead of via the backhaul link to get resources from core network.Cache technology can effectively reduce the backhaul congestion and improve the quality of user experience,so we will study the performance of the cache-enabled network in this thesis.At first,we study the performance analysis of multi-antenna multi-user heterogeneous network with random caching strategy.Compared with the existing research results,the system model we adopt is more general,and it includes some special scenarios,such as single antenna system and homogenous network.However,the complex system model leads to the complexity of derivation process,the main difficulty of our work is the n-th derivative of Laplace transform,which is caused by the multi-path effect of multi-antenna transmission.We use poisson point process to model the deployment of nodes in system,and utilize stochastic geometry,especially the recurrence between the n-th derivatives of the Laplace transform,to get the accurate expression of success probability,as well as the upper bound expression and the asymptotic expression,they are all functions about cache distribution of files.Then,according to the expressions and the simulation results,we can analyze the effects of various parameters on the system performance,and provide valuable advice for reasonable design of communication system.Due to the random cache strategy,the base station that is closer to users but does not cache the files which required by users will cause interference to users.In order to eliminate the interference caused by base stations,we adopt a content-aware interference nulling technology which controls the number of base stations that need to mitigate interference by the interference nulling coefficient and the distance between the user and its home base station.In addition,the size of the interference nulling coefficient and the distance between user and its home base station depend on the file requested by the user.Compared to interference nulling technology with fixed number of base stations to mitigate interference,the interference nulling technology adopted in this thesis is more flexible,it can reduce the interference,as well as ensure that the space freedom of the base station is not excessively used.Based on stochastic geometry,we get the closed-form expression and upper bound expression of the performance metric.Simulation results show that,compared to the random cache network without interference nulling,the performance of the system can be improved greatly by the content-aware interference nulling strategy that we adopted.