Research On Channel Estimation Algorithms For Massive MIMO System

Massive input multiple output(MIMO)technology is considered to be one of the key technologies in fifth-generation mobile communication(Fifth Generation,5G)systems due to its higher transmission rate and spectral efficiency.In order to take advantage of its many potential advantages,it is crucial to accurately obtain the channel status information(CSI)of the base station.However,in a massive MIMO system,hundreds of the antennas configured at the base station is getting massive,and the number of mobile users is also increasing rapidly,which makes the pilot frequency required by the base station too large,and the traditional channel estimation method has shortcomings such as inaccuracy in obtaining channel state information.Therefore,this study suggests two updated channel estimation algorithms to increase the channel estimation accuracy and reduce the overhead of the number of pilots by using the inherent low rank of the massive MIMO channel matrix based on the matrix completion theory and the proximal point.The specific research contents of this paper are as follows:Massive MIMO systems often use frequency division duplex(FDD)and time division duplexing(TDD)working modes.In the large scale Multiple-Input multiple output(MIMO)system,to reduce the high pilot overhead and improve the low accuracy of channel estimation in Frequency division duplex(FDD)mode,an adaptive step size gradient channel estimation algorithm is proposed in this paper.By using massive MIMO channel characteristics,the channel estimation problem is transformed into a low-rank matrix optimization problem,and an adaptive two-point step gradient descent method(SVP-BB)with singular value projection is proposed to recover the channel state information.Compared with gradient descent method,Newton method,and singular value projection based hybrid low-rank matrix recovery algorithm(SVP-H),the SVP-BB algorithm can obtain a lower normalized mean square error(NMSE)value in the first five steps of calculation.Simulation results show that the proposed algorithm has higher accuracy and better robustness in channel estimation.In a multi-user massive MIMO system,the number of antennas at the base station and the number of single-antenna users in the cell are much higher than the number of scatterers in the environment.If the scatterers are limited and all users share the same angle of arrivals,then the correlation between the channel vector increases,and the high-dimensional channel matrix can be approximated as a low-rank matrix.Aiming at the problem of unsatisfactory channel estimation accuracy and high pilot overhead of uplink channel in multi-user massive MIMO system in TDD mode,this paper proposes a new channel estimation algorithm by taking advantage of the low rank of massive MIMO channel matrix.By transforming the channel estimation problem into a low-rank moment recovery problem,a non-convex relaxation optimization model is established,and to improve the estimation performance,we recover the CSI using a proximal point algorithm.The simulation results show that for various finite scatterers under different signal-to-noise ratios,Compared with iterative weighted singular value threshold algorithm,fast iterative singular value threshold method and fast iterative weighted singular value threshold method,the proximal point algorithm obtains a lower NMSE value at the beginning,and has better convergence while effectively reducing pilot overhead.