Modeling And Performance Analysis Of Heterogeneous Cellular Networks Based On User Cluster Distribution

Ultra-dense heterogeneous cellular network is one of the key technologies for the successful operation of the fifth-generation mobile communication(5G).Uniformly deployed macro base stations provide most area coverage,while small base stations are used to supplement channel capacity in hot spots and fill blind areas in ultra-dense heterogeneous cellular networks.In order to meet the needs of users,the deployment density of low-power small base stations will be more than 10 times that of existing base stations in the future.Therefore,modeling and performance analysis of base station and user distribution in ultra-dense heterogeneous cellular networks efficiently and accurately is an urgent problem to be solved.Because dense small base stations are closely related to the spatial location of users,the cluster distribution of cellular users centered on small base stations should be considered in the analysis of network performance,thus reflecting the spatial location relationship between small base stations and users.This paper focuses on heterogeneous cellular networks,Base stations are distributed as Poisson Point Process(PPP)and users are distributed as Poisson Cluster Process(PCP)with small base stations as parent nodes.According to stochastic geometry theory,the theoretical expressions of coverage probability and energy efficiency are deduced,and the factors affecting network performance are analyzed through simulation experiments.The main research work of this paper is divided into four aspects:1.Modeling and coverage probability analysis of four cases of PPP and PCP distribution in two-dimensional space for base stations and users in heterogeneous cellular networks.First,under the Rayleigh fading channel,the signal to interference ratio(SIR)obtained by the user receiving signals from each layer of base stations is calculated based on the four models.Secondly,according to the cell access criterion of the maximum signal-to-interference ratio and the probability generation function of the Poisson point process and the Poisson cluster process,the sum-product function of the different layer base stations is calculated,and the coverage probability expressions of the four models are derived.Finally,the coverage probability of the four models is compared by Monte Carlo simulation experiments.The experimental results show that the base station distribution is PPP,and the user distribution is the best performance of the PCP model with the small base station as the parent node.Therefore,the follow-up work of this paper mainly focuses on the performance research and analysis of this model.2.Aiming at the two-dimensional spatial distribution model of Poisson clusters with small base stations as parent nodes in heterogeneous cellular networks,the effects of SIR threshold,Matern cluster process radius,Thomas cluster process variance,small base station density ratio and power ratio on coverage probability are analyzed by simulation experiments.3.The theoretical expression of energy efficiency is derived for the two-dimensional spatial distribution model of Thomas cluster process in which the user uses the small base station as the parent node in the heterogeneous cellular network.Firstly,the cluster of the target user is defined as the layer 0 base station.Secondly,according to the characteristics of the Thomas cluster process and the Poisson point process,the distance function and the association probability of the user accessing the different layer base stations are obtained,and obtaining the density of the layer 0 base station from the associated probability expression,thereby deriving an energy efficiency expression.Finally,the effects of variance and SIR threshold on the energy efficiency of Thomas cluster process are analyzed by simulation experiments.The experimental results show that the larger the variance of the Thomas cluster process,the thinner the user cluster and the lower the energy efficiency.4.Aiming at the dense urban environment of business office area,high-rise residential area and large shopping center,a three-dimensional spatial model based on Poisson Cluster Process is proposed and the network coverage probability is analyzed and simulated.The model assumes that the distribution of small base stations is a three-dimensional Poisson point process,and the user distribution is a three-dimensional PCP centered on small base stations.The distribution of sub-nodes is considered as Matern cluster process and Thomas cluster process.Based on this model,firstly,the cluster center base station of the target user is defined as the 0-tier base station,and then the SIR and interference characteristics of the 0-tier base station are analyzed according to the normal distribution and uniform distribution characteristics.Secondly,under Rayleigh fading channel,the coverage probability expression of the three-dimensional heterogeneous cellular network is derived.Finally,the theoretical conclusion is verified by Monte Carlo simulation experiments.As a result,the influence of SIR threshold,radius of user cluster,density ratio of small base station and power ratio on coverage probability is analyzed,and the network performance of three-dimensional user cluster distribution model,three-dimensional user uniform distribution model and two-dimensional user cluster distribution model is compared.The experimental results show that heterogeneous cellular networks based on three-dimensional user cluster distribution model have better coverage probability and more accurate coverage probability analysis.