Research On Low Complexity Hybrid Beamforming In Uplink Mmwave Massive MIMO Systems

Millimeter-wave(mmWave)massive multiple-input multiple-output(MIMO)systems have become one of the key technologies of the fifth generation of mobile communication systems(5G)because of its abundant available spectrum resources and spatial degrees of freedom.The short wavelength of mmWave facilitates shorten the distance between antennas so that the base station can be deployed with hundreds or thousands of antennas to form a large-scale antenna array.At the same time,massive MIMO can use the array gain to significantly improve the coverage and compensate for propagation loss.Therefore,the combination of mmWave and massive MIMO technology occupies an important position in enhancing 5G mobile broadband services.The traditional full-digital beamforming scheme requires a specific radio frequency(RF)chain for each antenna.The disadvantage of using this scheme in mmWave massive MIMO system is high hardware cost and power consumption.Therefore,using a hybrid beamforming scheme that combines digital and analog beamforming only requires much fewer RF chains than antennas,which can significantly reduce power consumption.However,the power consumption of the hybrid beamforming architecture system is still high due to the use of high-resolution analog-todigital converters(ADCs).Moreover,the hybrid scheme needs to jointly design the digital and analog beamforming,so the computational complexity is high.Regarding the high power consumption of mmWave massive MIMO systems,this dissertation considers low-resolution ADC mmWave massive MIMO systems and compared the achievable rates and energy efficiency of full-digital and hybrid beamforming.The uplink signal model of the low-resolution ADC mmWave massive MIMO systems is established according to the additive quantization noise model(AQNM).On this basis,this dissertation analyzes the influence of quantization bits on the achievable rates and energy efficiency of the hybrid and fulldigital beamforming scheme.This dissertation comprehensively compares the achievable rates and energy efficiency of mmWave massive MIMO systems under hybrid and full-digital architectures.It also provides the sufficient condition that the hybrid beamforming is better than the full-digital beamforming.Simulation results show that high quantization accuracy is not necessary for mmWave massive MIMO systems,and a hybrid beamforming architecture with low-resolution ADC can balance the achievable rates and power consumption.Regarding high signal processing complexity of mmWave massive MIMO systems,the lowcomplexity hybrid beamforming scheme is studied in this dissertation.In order to ensure the energy efficiency of the system and reduce the computational complexity of signal processing,a detection algorithm with low-complexity angular-domain compressing is proposed.By using the independent multipath channel model,the signal processing can be carried out from the antennadomain to the angular-domain.The angles with the strongest signal power can be selected to design the analog beamforming matrix.In digital baseband processing,the performance of the proposed scheme with zero forcing(ZF),maximum ratio combining(MRC)and minimum mean square error(MMSE)are compared.Simulation results show that compared with the existing angulardomain hybrid beamforming algorithms,the proposed scheme has higher achievable rates and significantly reduces computational complexity.