Performance Analysis For Intelligent Reflecting Surface-Assisted Heterogeneous Cellular Networks

Intelligent reflecting surface(IRS)assisted wireless communication technology has become a research hotspot in the field of communication as a promising research direction.By integrating a large number of low-cost passive reflection elements and regulating the phase of the reflection elements,IRS can intelligently configure the wireless propagation environment and significantly improve the wireless network coverage,spectral efficiency,energy efficiency and other performance indicators with low deployment cost and energy consumption.At present,most of the work focuses on the analysis and optimization design of IRS-assisted small-scale or single-tier cellular networks,and there is little research on IRS-assisted Heterogeneous Cellular Network(HCN).Due to the complexity of HCN interference,there is not only inter-tier interference between macrocells and small cells,but also intra-tier interference between macrocells/small cells,and the deployment of IRS will bring additional reflection interference.The heterogeneous characteristics of the network and the spatial random distribution characteristics of the interference nodes further increase the complexity of network interference,and the addition of IRS will change the propagation characteristics of wireless signals,affecting the distribution of useful signal power and network interference.These new changes have further increased the difficulty of evaluating network performance,and it is urgent to build a new theoretical analysis model of network performance.Therefore,how to effectively evaluate the network performance of IRS-assisted HCN,so as to quantitatively describe the gain of IRS to network performance and guide the deployment of IRS in HCN is an important research direction.The main work of this dissertation can be briefly described as follows:(1)In view of the problem that IRS-assisted HCN interference is complex and network performance is difficult to evaluate,a network performance analysis model was constructed by using the stochastic geometry theory,and the gain of IRS on network performance was studied.Firstly,beamforming technology is used to process the reflected signal,and a phase shift matrix of IRS associated with typical users is designed,so as to realize the in-phase superposition of direct signal and reflected signal.Secondly,the cascaded channel after the introduction of IRS is modeled,and the desired signal power distribution and interference signal power distribution at typical users are solved.Finally,the expressions of network coverage and spatial throughput are derived,and the influence of IRS deployment density and number of reflective elements on network performance gain is quantified.Experimental results show that through reasonable beamforming design,IRS can significantly improve the coverage and throughput performance of single-antenna HCN,compared with traditional HCN,the network coverage can be increased by up to 35%,and the spatial throughput can be increased by 60%.(2)In view of the problem that the precoding of base stations affects the design of IRS phase shift matrix under multi-antenna configuration,three different design schemes of base station precoding matrix and IRS phase shift matrix are considered,and the gain of IRS on network performance under different schemes is quantitatively analyzed.Firstly,for the active beamforming design on the base station side and the passive beamforming design on the IRS side,three different design schemes are considered under the condition that different channel state information(CSI)is known.Secondly,the cascade channel after the introduction of IRS in the multi-antenna scenario is modeled,and the desired signal power distribution and interference signal power distribution at typical users are solved under three different design schemes.Finally,the expressions of network coverage and spatial throughput are derived,the performance analysis theoretical model of IRS-assisted multi-antenna HCN is constructed,and the influence of different precoding matrix design schemes,the number of base station antennas and other parameters on network performance gain is quantitatively analyzed.Experimental results show that the precoding matrix of BS and the phase shift matrix of IRS can greatly improve the network coverage and spatial throughput performance by reasonable design.Compared with traditional multi-antenna HCN,the coverage can be increased by up to80%,and the spatial throughput can be increased by 270%.(3)In view of the problem that the limited deployment cost in the actual network,the impact of IRS deployment density on network spatial throughput is studied by considering the single-antenna and multi-antenna antenna configurations,and the network performance analysis theoretical model is used to study the impact of IRS deployment density on network spatial throughput.Given the total deployment cost of the small cell tier and the IRS tier,the optimal deployment density ratio of the IRS tier and the small cell tier is determined with the goal of maximizing the network spatial throughput.The experimental results show that the optimal deployment density ratio between the IRS tier and the small base station tier is 1 in the single-antenna configuration and 2 in the multi-antenna configuration.