Research On User Association And Backhaul Bandwidth Allocation For 5G Heterogeneous Networks

With the development of the Internet of Things,traditional communication technologies no longer meet the growing demand of mobile data traffic,which drives the development of the 5th generation mobile communication(5G).In order to meet the requirements of the 5G application scenarios,5G networks use the millimeter-wave technology,ultra-dense network,multi-network convergence,large-scale antenna arrays and so on.In the 5G ultra-dense heterogenous network(HetNet),the wireless access and wireless backhaul technology not only reduces network deployment costs and improves the flexibility of networking,but also facilitates massive devices to access the network.Considering channel variation and user mobility,the system needs to properly select the serving base station according to the topology of base stations and users,and effectively allocate resources according to the current loads of base stations.Therefore,a reasonable and effective user association and wireless bandwidth resource allocation scheme for 5G wireless HetNet is indeed needed.This paper focuses on the user association and wireless backhaul bandwidth resource allocation for 5G wireless HetNet.Firstly,it constructs a two-layer HetNet system model.And the zero-forcing beamforming technology is used to derive the channel capacity formulation.In addition,reverse time division duplex technology and base station sleeping technology are both used to perform interference management in the wireless HetNet.Then based on the two-layer HetNet system model,this paper establishes a mathematical model for the optimization problem 1(PI)of user association and backhaul bandwidth resource allocation in the 3.5 GHz low frequency system.And establishes a mathematical model for the optimization problem 2(P2)of user association and backhaul bandwidth resource allocation in the high frequency system which uses the millimeter wave band and low frequency band(Milimiter Wave + Low Frequency,mmWave+LF).Finally,a joint optimization algorithm is proposed for the P1 problem of low-frequency systems,and a greedy algorithm is proposed for the P2 problem of high-frequency systems.In this paper,the 3GPP(the 3rd Generation Partner Project)channel modeling based random distribution scene and the Ray Tracer(RT)application scenario are used to verify and evaluate the low-frequency system joint optimization algorithm and the high-frequency system greedy algorithm.Three traditional user association and resource allocation algorithms are used as comparison algorithms.The experimental results show that the performance of the joint optimization algorithm is optimal in the low-frequency system simulation.The edge user rate obtained by the joint optimization algorithm is at least 4%higher than that of the traditional algorithms,and the average user rate obtained by the joint optimization algorithm is at least 10%higher than that of the traditional algorithms.In the high-frequency system simulation,the greedy algorithm proposed in this paper has the best performance.The edge user rate obtained by the greedy algorithm is at least 3.21%higher than that of the traditional algorithms,and the average user rate obtained by the greedy algorithm is at least 89%higher than that of the traditional algorithms.