二维快速子空间DOA估计算法  

A fast subspace algorithm for two-dimensional DOA estimation

作　　者：马洪[1] 杨琳琳[1] 黎英云[1] 

机构地区：[1]华中科技大学电子与信息工程系武汉光电国家实验室(筹),湖北武汉430074

出　　处：《华中科技大学学报（自然科学版）》2008年第4期20-23,共4页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：航天支撑技术基金资助项目(20060307)

摘　　要：提出一种二维快速子空间DOA估计算法,该算法利用阵列协方差矩阵的一个子矩阵得到降维的信号子空间,不需估计整个阵列的协方差矩阵,也不需进行特征值分解,从而使得该方法具有运算量小、复杂度低和易于实时处理的特点,因而可以应用在小数据样本和快速时变的信号环境中.理论分析和计算机仿真结果表明:与MUSIC算法相比,该算法运算量最多为MUSIC算法的1/4,低信噪比条件下DOA估计性能损失并不大,当信噪比大于5dB时,性能与MUSIC算法相当.A fast subspace algorithm for two-dimensional DOA （direction of arrival） estimation is presented, in which the signal subspace of reduced dimension from a submatrix of the array covariance matrix can be obtained without eigen-decomposition. It only needs to estimate the subarray covariance matrix instead of the whole array covariance matrix, thereby indicating that the algorithm is computationally advantageous and more easily realization. The algorithm is suitable for application in shorttime sampled data and fast time-varying environments. Theoretical analysis and computer simulation results show that the computation load of the new algorithm is one fourth of that of the basic MUSIC algorithm. Compare to MUSIC, there is slight degradation of performance in low SNR （signal-to-noise ratio）, having similar performance when SNR is more than 5 dB.

关 键 词：阵列信号处理 DOA(波达方向)估计 快速子空间算法 谱峰搜索 计算复杂度 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]