The Research On The Key Technology Of Hybrid Beamforming In Millimeter Wave Massive MIMO System

Massive Multiple-Input-Multiple-Output technology(MIMO),as the evolution of traditional MIMO technology,dramatically improves the the multiplexing and diversity gain of the wireless communication system by increasing the number of antennas.Facing the increasingly limited spectrum resource problem,the research and use of millimeter wave(mm Wave)frequency bands have become a hot topic in both academic and industrial worlds.Based on the combination of the above two technologies,this thesis investigates the hybrid beamforming technology which can significantly reduce the system hardware complexity and provide beams with high directivity and gain in order to improve the reliability.Firstly,this thesis illustrates the composition and model of the massive MIMO system in detail,using the Saleh-Valenzuela method to build the mm Wave channel.Then,this thesis deeply investigates the hybrid beamforming algorithm based on beamspace,orthogonal matching pursuit and phase extraction respectively.According to the spatial sparsity of mm Wave channel,the aforementioned beamspace algorithm performs beamforming in the beam domain,while the orthogonal matching pursuit method transforms the beamforming design into a sparse signal reconstruction.The phase extraction approach avoids a large amount of iterations through extracting the phase of the optimal beamforming matrix.This thesis also compares the performance of these algorithms by simulation and comprehensively analyses their advantages and limitations.Afterwards,considering that traditional methods for hybrid beamforming are merely based on the channel matrix,this thesis innovatively proposes a symbol-level hybrid beamforming algorithm which combines the channel matrix and transmit signals jointly for hybrid beamforming realization.The simulation results show that the performance of the proposed algorithm is significantly improved with good scalability for multi-stream and multi-user transmission,whereas it requires higher refreshing rate for hardware devices.In order to find the best applicable scenario of the above algorithm to minimize the impact of its shortcomings,this thesis creatively develops a channel estimation algorithm with uplink pilots under a hybrid architecture to solve the problem that it is difficult for the base station to estimate the channel through traditional pilot methods under the hybrid architecture.Finally,this thesis proposes an energy efficient antenna selection assisted hybrid beamforming algorithm.The conventional antenna selection algorithms with Branch and Bound,maximum gain criterion and incremental or decremental search are firstly surveyed.Then,a Minimum Transmit Power(MTP)criterion whose aim is to minimize the transmit power while the same performance as the optimal beamforming algorithm is achieved under the antenna selection assisted hybrid beamforming architecture.Subsequently,with the proposed MTP criterion,this thesis further develops a novel antenna selection assisted hybrid beamforming algorithm and the simulation results show that the proposed algorithm has advantages in both spectrum efficiency and energy efficiency.