Research On Parameter Estimation Method Of Wireless Channel And Signal Detection Algorithm In Massive MIMO

As a new mobile communication system,5G has become a global research hotspot in recent years,in which the Massive MIMO as one of the key technologies has the advantages of high spectral efficiency and high channel capacity.But the research on this system is still not enough.Aiming at this issue,this paper draws up from two aspects,one of them is to deepen the understanding of the wireless channel through the parameter estimation method of wireless channel,the other is to analyze the performance of the detection algorithm and improve it for its shortcomings.In the studying of the parameter estimation method of wireless channel,this paper analyzes the principle of SAGE algorithm based on the introduction of ML algorithm and EM algorithm principle.To reduce the slow convergence rate of the algorithm caused by the accumulation of errors existed in continuous interference cancellation,an improved SAGE algorithm with partial interference cancellation as its initial method is proposed in this paper.The partial interference cancellation algorithm used in the proposed algorithm is to eliminate the error accumulation caused by continuous elimination in the classical SAGE algorithm by introducing an interference factors with "estimation ? recovery ? cancellation" as its main idea.The simulation results show that,the SAGE algorithm can accomplish the parameter estimation with higher accuracy,and the proposed algorithm can accelerate the iterative convergence speed of the algorithm on the basis of achieving the accuracy of the classical algorithm.To sum up,the proposed algorithm enhances the performance of the classical SAGE algorithm.In the studying of signal detection algorithms in Massive MIMO,this paper analyzes two kinds of linear detection algorithms(ZF detection technology and MMSE detection technology)and three kinds of nonlinear detection algorithms(ML detection technology,ZF-SIC detection technology and MMSE-SIC detection technology).The results show that the nonlinear detection algorithm is slightly higher than the linear detection algorithm in the detection accuracy,but its computational complexity is very high.Compared the advantages and disadvantages of these five algorithms,this paper concludes that the most prominent comprehensive performance of these five algorithms is MMSE algorithm.In order to further improve the performance of the MMSE algorithm,this paper proposes a MMSE-BI algorithm based on block iteration.The block iterative algorithm avoids the problem of large matrix inversion in the classical MMSE detection algorithm in the form of iteration,which greatly reduces the computational complexity of the algorithm.The simulation results show that the proposed algorithm has lower complexity under the premise of achieving the near-optimal performance of the classical MMSE algorithm.Compared with the MMSE algorithm,the proposed algorithm has more advantages in performance.