Research On Channel Estimation And Hybrid Precoding Techniques For Millimeter Wave Systems

Due to the large available bandwidth and higher spectral efficiency,the integration of large-scale multiple input,multiple output(MIMO)and millimeter wave(mmWave)technologies can achieve an order-of-magnitude increase in system throughput,making them future 5G wireless communication systems.The key technology.Millimeter Wave Transmission Using a large antenna array at the transmitter and receiver,combined with millimeter wave frequencies and mixed signal power limitations,new mass MIMO signal processing techniques are needed.Due to the large bandwidth,low-complexity transceiver algorithms have become very important.It has become very significant to study the channel estimation techniques and hybrid precoding techniques for large-scale MIMO millimeter-wave systems.This paper focuses on the 5G physical layer technology under the condition of millimeter wave massive MIMO.The main contents include the study of large-scale MIMO millimeter-wave channel estimation techniques to obtain millimeter-wave channel state information(AoA/AoD and channel gain);and study the hybrid precoding technology of millimeter-wave systems to improve the channel capacity and performance of the system;Single-user hybrid precoding is extended to multi-user mixed precoding,which is more in line with current application prospects.The main work is as follows:(1)In large-scale MIMO millimeter-wave systems,how to accurately estimate channel state information(CSI),which is crucial for precoder and combiner design,is a challenging problem.By exploiting the sparseness of millimeter-wave channels,a low-complexity algorithm for estimating millimeter-wave signal parameters is proposed in this paper.It can more accurately estimate the millimeter-wave channel state information AoD/AoA and channel gain,which not only overcomes the analog Beamforming hardware constraints also enable near-digital solution performance.(2)Due to hardware costs and power limitations in millimeter-wave systems,it is difficult to achieve full digital precoding of large antenna arrays.The solution is usually to use hybrid precoding techniques.Based on the quantized variable phase shifter's analog beamforming and the use of a limited number of RF chains,we propose an improved OMP-based optimization algorithm for largescale MIMO millimeter-wave systems without adding complexity.,can achieve better performance than OMP.(3)Multiuser hybrid precoding can be used to further increase the spectral efficiency of millimeter wave systems.In this paper we study the hybrid precoding of multiuser millimeter-wave systems and propose an optimized hybrid minimum mean square error(MMSE)precoder.The simulation results show that the proposed precoder is relative to the known two-stage.Hybrid ZF precoders have significant performance advantages.