Research On Wireless Channel Modeling And Channel State Information Acquisition Of Massive MIMO Systems

The rapid development of the Internet of Things(Io T)and mobile Internet has led to the transformation and upgrading of a new generation of wireless communication networks.On the one hand,the wireless signal transmission quality is mainly dependent on its physical channel characteristic.Developing the channel model that is capable of accurately abstracting wireless communication environment is the important link of performance analysis and design optimization for wireless communication systems.Unfortunately,the traditional channel models designed mostly for cellular networks have their own advantages and disadvantages in terms of accuracy,complexity and portability.They lack of effective means to accurately characterize the characteristics of the diversified scattering scene and have the weak compatibility with the thing-to-thing wireless communication channel in the mobile Io T scenario.On the other hand,massive multiple-input multiple-out(MIMO),as one of the transformative technologies in the physical layer of the 5th Generation Mobile Networks(5G),can deeply excavate the degree of sptial freedom by deploying a larger number of antenna arrays.This promising technology leads to great potential to further improve the spectrum efficiency,energy efficiency and robustness of the communication systems.The acquirement of accurate channel state information(CSI)is the key to achieve the performance advantages of massive MIMO.However,A series of problems,such as pilot contamination in time division duplexing(TDD)mode for both cellular networks and new cell-free networks and excessive channel feedback overhead caused by antenna scale expansion in FDD mode,seriously affect the acquisition process of CSI and restrict the smooth deployment and compatible implementation of massive MIMO systems.Focusing on the above-mentioned urgent problems,this paper conducts an indepth research on the key technologies of wireless channel modeling and channel state information acquisition for massive MIMO systems.The research content and detailed work are as follows:(1)Research on multi-antenna channel modeling for mobile Internet of Things(Io T)scenarios.Focusing on the differentiated characteristics of the mobile Io T scenarios and comprehensively considering the accuracy,complexity and adaptability of the channel model,a novel three-dimensional multi-antenna channel model is proposed in this paper by adopting the geometric-based stochastic channel modeling method.Specifically,the proposed channel model combines the characteristics of antenna deployment and scatterer distribution,and introduces a three-dimensional double-spheres model and an ellipsoid model to characterize the multipath effects produced by the local and remote scatterers,respectively.According to the theoretical model,we have derived the corresponding spatial-temporal correlation characteristics and spatial-Doppler power spectral density and have analyzed the combined effect of model parameters on channel correlation.Finally,an effective parameter calculation method is utilized to develop a corresponding simulation model.It greatly improves the efficiency of theoretical analysis and simulation for multi-antenna wireless communication channel.(2)Research on alleviating the pilot contamination for TDD cellular massive MIMO networks.Pilot contamination is regarded as the main factor restricting the performance of TDD massive MIMO systems.Combining the advantages of fractional pilot reuse and pilot assignment optimization,this paper proposes a fractional pilot reuse and max k-cut based pilot assignment scheme.Firstly,the metric of the user's susceptibility to interference is designed,and the traditional fractional pilot reuse scheme is modified by considering the cell factor as the basis for dividing the center-edge boundary.Secondly,the pilot assignment process is innovatively regarded as the cutting of the graph,and the pilot assignment optimization is mapped to the solution of the max k-cut problem in graph theory.The simulation results show that the proposed scheme can significantly reduce the global pilot contamination and enhance the fairness of the communication service quality for the cellular massive MIMO systems.(3)Research on alleviating the pilot contamination for TDD cell-free massive MIMO networks.The cell-free networks have the ability to solve the inherent disadvantages of the traditional cellular paradigm,such as intra-cell differential data rate variations and inter-cell interference.In this paper,we have derived the expressions of the uplink and downlink spectral efficiency of the cell-free massive MIMO systems,and have redefined the indicators for the severity of potential pilot contamination among users for cell-free network topology.In this way,a dynamic weighted pilot contamination graph is constructed and its solution of max k-cut is obtained to achieve global pilot assignment optimization.In addition,the pilot power control coefficients are introduced and a sparrow search algorithm-based joint pilot-data power control scheme is proposed.By leveraging the Logistic chaotic mapping and adaptive t-distribution variation strategy,the optimal solution search ability and convergence speed of the proposed scheme are improved.(4)Research on compressed channel feedback for FDD massive MIMO Systems.The costly feedback overhead generated by large-scale arrays poses an arduous challenge to the traditional channel feedback scheme.In this paper,a sparsity learning-based CSI feedback method in a compressive sensing framework for massive MIMO systems is proposed.The key insights are to learn the sparsity structure of CSI through the recursive least squares algorithm and to process a continuous update of the sparse basis.Accordingly,a CSI sparsity characteristic adaptive dictionary is constructed.In addition,an adaptive forgetting factor is introduced to reduce the dependence of initialization with improved convergence gradually.The simulation results reveal that the proposed scheme achieves excellent performance in terms of both compression efficacy and recovery accuracy.