Research On Beam Selection And Channel Estimation In Millimeter Wave Massive MIMO System

With the continuous improvement and progress of mobile communication technology,the way of people's life has been completely changed and the development of related industries also have a further promote.At present,people's demand for data transmission rate,system capacity,data throughput etc.in the wireless communication system is increasing.The fifth generation mobile communication system that meets the higher requirements of user experience has aroused great interest in the researchers.As one of the key technologies of 5G,Massive MIMO technology has become a hot research topic in recent years.As the frequency spectrum of the low frequency section is very crowded,the abundant available spectrum resource of millimeter wave catches the people's attention.The perfect combination of millimeter wave and Massive MIMO Technology brings abundant bandwidth resources,higher data transmission rate and energy efficiency and has already become another research hotspot in the field of communication.Therefore,this paper mainly analyzes and studies the related scheme of beam selection and channel estimation for millimeter wave Massive MIMO system.At the first chapter of this paper,it introduces several classical basic models of Massive MIMO system,and the analysis of the related characteristics under different models is also given at the same time.Then several common linear precoding schemes used to deal with the sending signals are clearly explained.The general scene of millimeter wave communication is also described,and a brief analysis of the sparsity of its channel is given.Due to the system channel state information is the key to the preprocessing of the signal,it is necessary to seek the channel estimation algorithm with higher accuracy and lower complexity.In this chapter,the classical channel estimation algorithm LS and MMSE based on pilot are introduced.Then to solve the problem of the recovery of sparse signal,the OMP and CoSaMP channel estimation algorithms based on compressed sensing technology are introduced.As the MMSE algorithm is too complex in the multi antenna system,only the other three algorithms are simulated and analyzed to better understand the performance of different algorithms.To solve the problem that high energy consumption of the traditional full digital precoding caused by large numbers of radio frequency chain in millimeter wave massive MIMO system,a beam selection scheme based on lens was proposed.Simulation results show that,the proposed beam selection can nearly reach the full performance of digital precoding and approach higher energy efficiency at the condition of reducing the number of the required RF chains and cut down the computation complexity effectively.Finally,with the problem that the ideal channel state information is difficult to obtain in the millimeter wave Massive MIMO system based on the beam selection scheme,combining the sparsity of the signal with the algorithm of the compressed sensing channel estimation,a block sparse regularization CoSaMP algorithm for millimeter wave Massive MIMO system is proposed to estimate the beam-space channel.The simulation results Indicate that contrast with the traditional CoSaMP algorithm,the proposed algorithm is more accurate in channel estimation and can also improve the energy efficiency of the system.