Low Complexicy Beam Training For Mm-wave Massive MIMO Systems

A key indicator of 5G is the transmission rate.According to the communications industry's expectation,5G should achieve a transmission rate of10 times faster than 4G,that is,The peak rate of 5G is not less than 10 Gbps.There are two ways to increase the transmission rate of wireless transmission,one is to increase the spectral efficiency,and the other is to increase the spectral bandwidth.Millimeter wave is a kind of electromagnetic wave with wavelengths of millimeter order.Its frequency is about 30 GHz~300 GHz.It has abundant spectrum bandwidth.And can meet the demand of 5G communications for spectrum resources.Millimeter wave is a kind of electromagnetic wave with wavelengths of millimeter order.It has abundant spectrum resources and bandwidth.It can meet the demand of 5G communication for spectrum resources.Millimeter wave is characterized by short wave length,short transmission distance,large transmission loss and weak penetration.On the other hand,the physical size of the antenna is proportional to the wavelength,and the millimeter wave allows both sides of the communications to be equipped with large-scale antenna arrays.The use of large-scale antenna array and beamforming technology can overcome the shortcomings of millimeter wave communications and improve the quality of communications.In millimeter-wave large-scale antenna array systems,in order to cope with severe path loss,beams at the transmitter and the receiver must be aligned very precisely.Combining with the sparse characteristics of millimeter-wave channels,if there is a slight beam misalignment between two connected devices,the signal power will be significantly reduced.Therefore,the beam training process must be used to determine the transmitter and receiver beamforming vectors.Considering the difficulty of full-digital beamforming and the constraints of analog beamforming,in this study,a hybrid analog-digital beamforming technique is used low-complexity beam training in millimeter-wave systems.The hybrid beamforming technology decomposes the full digital processing into two parts: digital baseband coding and analog Radio Frequency beamforming,which can reduce the processing complexity and hardware requirements while obtaining a higher gain.The hybrid transceiver also has the ability of simultaneous multi-directional scanning.The details are as follows:According to the characteristics of the array antennas,a hybrid beamforming codebook set is defined to cover all spatial angles,and the transceiver codebook is optimized based on a certain optimal criterion.The best transmitting codebook andreceiving codebook are determined by a beam search algorithm,and the corresponding spatial angles of transmitting codebook and receiving codebook,namely angle of arrival and angle of departure,are then obtained.In this paper,an improved hierarchical beam search method is proposed by optimizing the hierarchical codebook.The simulation results show that the success rate of beam search based on the optimized codebook is improved by about 10%.The millimeter wave channel restricts scattering,and the multipath components are mainly generated by reflection.Different multipath components have different physical transmission angles and receiving angles.In order to cope with the failure of the beam link caused by the user's movement or rotation,a dual-beam search scheme is examined in this thesis.The optimal beam and the sub-optimal beam are obtained simultaneously.The optimal beam is used for communications,and the sub-optimal beam is used as an alternative beam to ensure the fast establishment of the communications link when the communications link fails.The simulation results show that the success rate of multi-beam search is not as high as that of single-beam search,and the success rate of sub-optimal beam search will be improved only if the optimal beam is searched accurately.