Channel Estimation For Massive MIMO-Based Millimeter-Wave Systems Based On Hierarchical Codebook

In recent years,the news about the 5th generation(5G)mobile communication system has emerged in an endless stream,which has aroused the attention of people from all walks of life.The commercial use of 5G has huge opportunities,all countries around the world are looking forward to getting a head start in this 5G race.After 5G enters commercial use,millimeter-wave communication is the main battlefield for 5G research and development.One of the noteworthy features of millimeter-wave signals is its severe path loss.Therefore,a feature of millimeter-wave communication is configuring the large-scale antenna array at the transmitter and the receiver to form a large-scale multiple input multiple output(MIMO)system and increase the array gain.With large-scale antenna arrays,millimeter-wave systems can use hybrid precoding at the transmitter and receiver to compensate the path loss caused by ultra-high frequencies,overcome additional noise and reduce out-of-cell interference.The hybrid precoding technique needs to design the precoding matrix and the combining matrix of the transmitter and the receiver according to channel state information(CSI).The acquisition of accurate CSI relies on channel estimation.Due to the transmission characteristics of millimeter waves,the channel estimation problem of a millimeter-wave massive MIMO communication system based on hybrid precoding is challenged.In this paper,two millimeter-wave massive MIMO channel estimation methods based on hierarchical codebook are studied:channel estimation method based on the codebook for the discrete Fourier transform(DFT)approach and channel estimation method based on the codebook for the beam widen(BW)approach,and then a modified hierarchical codebook scheme is proposed based on the BW codebook.The method of code word design is studied for the DFT codebook and the BW codebook,and the corresponding channel estimation algorithm is given for the improved BW codebook.Through simulation,the system performance of the three codebooks is compared from the four aspects of time complexity,beam model,search success rate and achievable rate.The simulation results show that the DFT codebook has better system performance,but at the cost of requiring more radio frequency(RF)links;the BW codebook has similar system performance with the DFT codebook with fewer RF links,and the modified BW codebook is more accurate in estimating the angle information in the channel,and has system performance close to or even beyond the DFT codebook.A millimeter-wave cellular network system model based on random geometry is studied.The signal to interference plus noise ratio(SINR)coverage and the achievable rate coverage of the millimeter-wave cellular network based on the system model are analyzed under the condition that the perfect CSI is known at the transmitter and the receiver.In order to be more realistic,the channel estimation method,the improved BW approach,for downlink point-to-point communication is extended to the millimeter-wave cellular network,the expressions of SINR and achievable rate of the millimeter-wave cellular network based on the estimated CSI is given.Simulations about the SINR coverage and achievable rate coverage of common and dense millimeter-wave cellular networks based on estimated CSI are analyzed.The simulation results show that the larger the average cell radius,the higher the channel estimation accuracy is;in the ultra-dense millimeter-wave cellular network,the larger the LOS path loss index is,the higher the channel estimation accuracy is.When the average cell radius is about half of the equivalent LOS area radius,the system can acquire the optimum SINR coverage.