Research Of Networking Strategy Combined With Caching And Mobility In Heterogeneous Ultra Dense Network

In recent years,with the rapid deployment of the 5th Generation Mobile Communication System(5G),more and more emerging mobile services are going to enter people's lives.Users also have higher requirements on performance such as network capacity,date rate and latency.The cache-enabled ultra-dense heterogeneous networking is considered to be one of the effective ways to increase capacity,increase date rate and reduce latency,but the dense deployment of base stations makes 5G networks more complex.The increasing mobility of users also makes the mobility management strategy more challenging.Therefore,this thesis focuses on the networking strategy that combines cache and mobility in dense heterogeneous network scenarios to provide users with better service quality.The main research contents of this article are as follows:Firstly,this thesis takes user mobility into the consideration in cache strategy research,and proposes a mobility-aware proactive cache strategy in dense heterogeneous network scenarios.Base stations are supposed to be able to perceive the user's mobile trajectory,the uncompleted download probability of users in the neighboring cell,and the content popularity of the neighboring base station.Then,base stations can modify the file popularity with them,thereby updating the cache content of the base station in advance.This process will improve the quality of cache service and reduce the average service delay.In addition,based on the layered architecture of the heterogeneous network,this thesis also designs a layered caching mechanism that combines user mobility with the network architecture to serve users with different mobility strengths.Files for different speed users are trend to be stored in different cache layer nodes,which result in improving the cache hit rate.Besides,this thesis formulate the cache placement matrix optimization problem with the aim of effective capacity and solved it by a genetic algorithm-based approach.Simulation results show that the proposed mobility-aware proactive caching strategy not only improves the system capacity and cache hit rate but also reduce the average user latency.Secondly,this thesis takes cache into the consideration of mobility management strategy research and proposes a dynamic cell range expansion(CRE)strategy in dense heterogeneous networks.The traditional mobility management strategy generally focus on the continuity of services,letting users switch to the base station with the best channel conditions as much as possible,which is without the consideration of delay.There are a large number of base stations in dense heterogeneous networks.Users accessing base stations that have cached the required content will obtain a lower service response delay.Based on the CRE technology,this thesis take cache into consideration and propose a dynamic CRE strategy.By dynamically adjusting the virtual range of small cells,we can control the handover timing and handover target of mobile users so as to improve the QoS of multiple services including low-latency services.This thesis constructs QoS evaluation utility functions for different services,and formulate the optimization problem.We introduce the Deep Q-Network algorithm to obtain the optimal CSO value strategy by training the agent to learn the network environment.Therefore,the virtual range of the cell can dynamically change as the user moves in different service scenarios.Simulation results show that the proposed strategy can significantly reduce the average delay of mobile users,improve system utility,and enhance the QoS guarantee capability of networks for multiple services.In summary,this thesis focuses on the requirements of services of 5G for speed and delay,and makes the research of networking strategies that combine cache and mobility in the heterogeneous ultra-dense network.Enhanced strategies that combine cache and user mobility are proposed,which improve the speed and delay performance of the network in multi-service scenarios and enhance the QoS guarantee capability of the mobile communication network.