Research Of The Spatial Correlation Of Wireless Multi-Access Channels Based On Fractal Theory

With the rapid of smart devices(smartphones,tablets,and portable computers),massive internet of things(Io T)devices,various new services and application scenarios,the traffic demand for wireless networks is rising rapidly.To meet this demand,the fifth generation(5G)mobile communication system proposes to combine the large-scale antenna technology,the millimeter wave technology,and the ultra-dense heterogeneous networks to increase network capacity.However,the wireless channel characteristics in 5G wireless communication systems will exhibit spatial differences which cannot be ignored.Due to the increase in user density,there is a partial overlap in the wireless transmission environments among users,the spatial correlation of wireless channels of adjacent users is enhanced.The spatial correlation of wireless multi-access channels is that the wireless channel states in the different directions of a base station have the certain correlation,and are no longer independent random variables.The current studies on wireless channel models paid less attention to the correlation among multiple channel states in different directions,which results in the multi-access channel model not being able to closely approximate the spatial variation law of actual channels in massive access scenarios.Irregularly things in nature generally have fractal characteristics.Moreover,the fractal theory is a discipline that is proposed to describe small-scale characteristics and the relationship between local and overall.The fractal theory is introduced to study the spatial change law of wireless multiaccess channels,especially the law of channel change for dense users in massive access scenarios,which is very effective and can solve the shortcomings of the existing wireless channel model in terms of spatial correlation.Therefore,the studying on the spatial characteristics of wireless channels is of great significance to the evolution of 5G communication technology.This thesis conducts in-depth study on the spatial characteristics of wireless channels.The main innovations are summarized as follows:Firstly,considering that wireless signals propagating in non-free space are affected by the environment and exhibit anisotropic fading,the coverage boundary of a cellular network,i.e.,a base station can cover in a specific direction with the maximum transmission distance,is used to characterize the impact of the propagation environment on wireless channels.To analyze the impact of anisotropic propagation environment on cellular coverage boundaries,the wireless channel state information is measured in real urban and suburban environments.The statistical characteristics of the wireless cellular coverage boundary are analyzed based on the fractal theory.The analysis results show that the wireless cellular coverage boundary has statistical fractal characteristics in the angular domain.Secondly,the ultra-dense heterogeneous network equipped with millimeter wave technology will become the major network architecture in the future.The coverage and handover performance are critical for small cell networks with the millimeter-wave transmission system.Based on the fractal characteristics of the cellular coverage boundary,a multi-directional path loss model is proposed to analyze the impact of anisotropic propagation environment on the performance of 5G small cell networks.Based on the proposed model,the coverage probability,association probability,and handover probability of the 5G small cell network are derived to investigate the impact of the anisotropic environment on cellular networks.Simulation results show that the access probability for short transmission distances(i.e.,50m)is significantly reduced with the increase of the anisotropic path loss effect in the 5G small cell network.The handoff rate in the anisotropic transmission environment is larger than that of the isotropic path loss model.Thirdly,based on the fractal characteristics of cellular network coverage,the theoretical analysis and experimental verification imply the long-range dependency feature existing in the coverage boundary on the angular scale.This paper proposes a fractal channel estimation scheme for massive multi-access scenarios by utilizing the spatial correlation of channels in different azimuths of a base station.With the aid of fractal theory and the measured coverage boundary data for base stations in Shanghai,China,a fractional autoregressive integration moving average model is built to describe the relationship among the maximum transmission distances of adjacent azimuths of the BS.Furthermore,the channel state information(CSI)of a channel is regarded as the function of the transmission distance and the maximum transmission distance at the azimuth of the channel based on the experience propagation model.Hence,the CSI of an undetected channel can be estimated based on the CSI of the adjacent detected channels.Based on the proposed fractal channel estimation scheme,the channel estimations of massive terminals can be carried out with a small number of pilot resources.Compared with the traditional least squares scheme,the pilot overhead of the newly proposed fractal channel estimation scheme is reduced by 94.34% when the terminal density is ten million per square kilometer.Fourthly,considering that the coverage boundary of wireless cellular networks has a fractal characteristic on the angular scale,it is realized that the longest distance covered by the base station in different directions has the self-similarity characteristic.Based on this result,the capacity domain of the wireless multi-access fractal channel is derived,and the bit error rate(BER)relationship model of two adjacent wireless channels is constructed.At the same time,a fractal channel coding scheme with rate adaptation is proposed to reduce the overall BER of the communication system while reducing the pilot overhead.Compared with the BER of fixed code rate schemes,simulation results show that the proposed fractal channel coding scheme can reduce the BER by 45.92% at most.In summary,this thesis aims at the studies on spatial correlation of wireless multi-access channels.Based on the fractal theory,the measurement experiments verify the fractal characteristics of wireless cellular network coverage.At the same time,the impact of fractal characteristics on the performance of millimeter-wave small cell networks is analyzed.Based on the self-similarity and long-range correlation of the coverage boundary of cellular networks,the fractal channel estimation and coding schemes are proposed to improve the performance of cellular networks.Compared with the traditional channel estimation scheme,the proposed fractal channel estimation scheme can reduce the pilot overhead.The study work of this thesis enriches researchers' understanding of wireless channel characteristics and provides a theoretical basis for the development of future communication technologies.