The Rerearch On Mode Selection And Resource Allocation Algorithm For Device-to-Device Communication In 5G Cellular Network

With the rapid growth of smart phones,as well as the explosive demand of mobile multimedia services,traditional cellular networks faces enormous challenges.For the future plans of 5G,due to the intensity of people in big cities,the communication distance between devices has been reduced,and new communication opportunities have emerged,D2D(Device-to-Device)has emerged as a candidate key technology for 5G.However,leading D2 D into traditional cellular networks will bring many technical problems.such as D2 D device discovery and link establishment,mode selection,resource allocation and mobility management.In order to solve the coexistence problem between them,this thesis will focus on the issue of D2 D mode selection and resource allocation underlying cellular networks.In most of the existing literature,D2 D communication mostly uses the traditional communication modes: cellular mode,D2 D shared mode,and D2 D dedicated mode.No more mode cases are considered and the system has not studied under the entire load system of the cell.In order to meet the needs of future 5G D2 D communication,this thesis studies the new D2 D communication mode: Local route mode,direct D2 D mode and relay-assisted D2 D mode,classifying the load status of base stations in single cell,and corresponding communication modes are designed for different levels.Based on the background,significance and current situation of D2 D communication research,this thesis classifying the network status of single cell base station into three level:Non-full load and full load.In the case of non-full load,it is subdivided into low load and medium load.Firstly,the interference of D2 D users reusing uplink resources of cellular users under various load conditions is analyzed in detail.Due to more free resources,this thesis propose a mode selection based on the maximum achievable rate and resource allocation scheme for low load.To deal with medium load scenarios,the number of free resources is relatively less,and the use of low-load solutions will lead to a sharp decline in the fairness of D2 D users.therefor,a mode preliminary scheme based on channel state is proposed,and maximizing the throughput of D2 D users through joint mode selection,power allocation,and resource allocation.Simulations show that the two schemes under non-full load can effectively improve the throughput of D2 D users on the basis of satisfying the fairness of cellular users and D2 D users.In addition,For the case of full load,due to no free resources under full load,so the non-full load option is no longer applicable.This thesis performes the calculation and analysis of power allocation,and proposing a scheme for joint mode selection and resource allocation based on Kuhn-Munkres algorithm.Simulation results show that comparing with the existing schemes,this scheme can effectively improvethe throughput of D2 D users and their reusable cellular users.