未知噪声背景下的改进型MMUSIC算法  

An Improved MMUSIC Algorithm in the Presence of Unknown Noise Field

作　　者：王科文[1] 谭鸽伟[1] 

机构地区：[1]华侨大学信息科学与工程学院,福建厦门361021

出　　处：《华侨大学学报（自然科学版）》2012年第4期388-391,共4页Journal of Huaqiao University(Natural Science)

摘　　要：针对不能确知被估计信号源的相关性和非平稳且噪声功率未知的高斯噪声环境,提出一种改进型的修正多重信号分类(MMUSIC)算法.用协方差差分法消除阵列协方差矩阵中的未知噪声矩阵,再用MMU-SIC算法对得到的协方差差分矩阵进行解相关或解相干处理.该算法只要求阵元数为不少于2L+1,即可在不影响相干信号源估计的同时,保证不相关信号源的估计性能.仿真结果表明:算法能够在不能确知信号源相关性和低信噪比的未知噪声环境下,对信号源进行有效的波达方向估计,获得很高的估计精度和角度分辨率.Aimed at the problems of unknown the correlation of signal sources and the gauss noise environment which is nonstationary and unknown noise power, an improved modified multiple signal classification （MMUSIC） algorithm is proposed in this paper. Firstly, covariance differencing approach is used to eliminate unknown noise matrix from array covariance matrix. Then, the covariance differencing matrix is decorrelated or decoherenced by MMUSC algorithm. The algorithm only requires that the number of array elements isn＇t less than 2L-＋1, it doesn＇t affect the estimation of coherent signal sources and simultaneously ensure the estimated performance of uncorrelated signal sources. Simulation results show that the method can effectively estimate the signals＇ direction of arrivat （DOA） in the presence of unknown correlation of signal sources and unknown noise filed with low signal to noise ratio （SNR）, and get very high estimation accuracy and angular resolution.

关 键 词：波达方向 未知噪声 修正多重信号分类算法 协方差差分法 

分 类 号：TN911.7[电子电信—通信与信息系统]