A Low Frequency Aided Precoder Design Of Massive MIMO Millimeter Wave Communication System

In order to meet the growing demand of wireless communication data transmission,millimeter wave has become the main research topic of communication technology because of its abundant spectrum resources and larger communication bandwidth.Considering to compensate for the high path loss in millimeter wave,large-scale antenna arrays are expected to be configured at the base station and the users' side to provide high beam gain and spatial multiplexing gain.For the purpose of improving the energy efficiency of the communication system,based on the the channel estimation information,a precoding scheme is deployed,so as to send the signal to the users' side more directionally and reduce the energy loss.However,due to the deployment of large-scale antenna arrays,the complexity of channel estimation on millimeter wave is greatly increased.Based on the spatial similarity between low frequency band and millimeter wave,a method to recover millimeter wave spatial information via low frequency information is proposed by researchers,which could speed up the process of channel estimation.Compared with the high cost of all-digital precoding,hybrid precoding can not only control the amplitude and phase,but also reduce the hardware overhead and achieve promising performance at a reasonable cost.In this paper,a low frequency aided precoding scheme is proposed for downlink multi-user massive MIMO millimeter wave hybrid precoding system.In most of the work on hybrid precoding architecture,multi-user interference is shelved in the analog domain but dealt with in the digital domain.In this paper,a low frequency aided millimeter wave massive MIMO analog precoding design and a robust analog precoding design are proposed,which can effectively suppress multi-user interference and optimize beamforming gain in the analog domain.The weighted sum algorithm is used to solve the multi-objective optimization problem of beamforming gain and multi-user interference,and the semidefinite programming algorithm is used to deal with the nonconvex constraint problem in the optimization process.Simulation results show that the performance of the analog precoding method proposed in this paper is better than that of the traditional analog precoding method.And in the case of angle estimation error,compared with the beam selection method,the robust analog precoding method proposed in this paper can improve the sum rate up to 110%.At present,there has been a research on low-frequency aided single-user millimeter wave beam selection.This paper extends it to study the low-frequency aided hybrid precoding design based on beam selection in multi-user scenarios.The zero-forcing algorithm is applied in the digital domain to eliminate multi-user interference.In the analog domain,the beamset that maximizes the system capacity is selected by beam search based on the codebook,which is pre-defined.On the basis of beam selection using low frequency information,five low frequency aided beam selection methods are proposed: the optimal low frequency beam set is obtained according to the low frequency information,and then the millimeter wave beam contained in the low frequency beam width is trained to obtain the optimal beam set of the millimeter wave system.The simulation results show that,compared with the in-band only beam selection,the low-frequency aided beam selection method can greatly reduce the training overhead and accelerate the millimeter wave beam selection process.