Research On Channel Estimation Method Of Millimeter Wave Massive MIMO System Based On Compressed Sensing Under Low-precision AD

Massive multiple input multiple output(MIMO)technology and millimeter wave(mm Wave)are the core key technologies in G wireless communications.However,millimeter wave massive MIMO systems are due to the large number of antennas and high frequency bands.Communication increases the amount of data processed by the receiving end exponentially,greatly increasing the energy consumption and hardware cost of the receiving end.In order to solve this problem,the academic circles proposed to adopt low-precision analog-to-digital converter(ADC)at the receiving end.Linear quantization error,which makes channel estimation more difficult.Traditional channel estimation algorithm does not consider the influence of quantization error,and the performance is severely degraded.This article focuses on the channel estimation problem of millimeter wave massive MIMO system with low-precision ADC..Three channels based on Quantizied Iterative Hard Thresholding(QIHT)and Hard Thresholding pursuit(HTP)are proposed for the channel estimation problem of millimeter-wave massive MIMO systems with low-precision(-bit)ADCs.The estimation algorithms are:(1)Newton-step-based Quantizied Iterative Hard Thresholding(NSQIHT).According to the optimization theory,Newton’s method contains more information than the fastest gradient descent algorithm.The NSQIHT algorithm uses Newton Method instead of the fastest gradient descent algorithm.(2)Quantizied Hard Thresholding pursuit(QHTP),based on the HTP algorithm,the QHTP algorithm takes into account the influence of quantization noise,and has better performance in the millimeter wave massive MIMO system in the low-precision ADC.(3)Newton-step-based Quantizied Hard Thresholding pursuit(NSQHTP).The NSQHTP algorithm uses Newton’s method to replace the fastest gradient descent algorithm based on the QHTP algorithm.Compared with the QHTP algorithm,It further accelerates the convergence speed and improves the convergence accuracy.For the channel estimation of millimeter wave massive MIMO system with-bit ADC,the optimal quantization threshold is considered,and a new channel estimation algorithm is proposed.Its contribution is summarized as:(1)It is found that the existence of the optimal quantization threshold point can improve the channel estimation performance,and the analysis shows that the optimal threshold achieves the optimal performance when it is infinitely close to the amplitude of the received signal.An adaptive Quantization Thresholding(ATQ)algorithm is proposed.To get the best quantization threshold.(2)The adaptive quantization threshold algorithm is combined with the NSQHTP algorithm,and the Adaptive Newton-Step-based Quantizied Hard Thresholding Pursuit(A-NSQHTP)algorithm is proposed.Simulation verification shows that the ANSQHTP algorithm has a certain performance improvement compared with the NSQHTP algorithm with a default fixed threshold and other mainstream algorithms.