基于子空间和投影分离的三维正弦信号频率估计算法  

3-D Frequency Estimation of Multiple Sinusoids Using Subspace and Projection Separation Approaches

作　　者：黄龙庭[1] 吴云韬[1] 廖桂生[2] 张彦铎[1] 

机构地区：[1]武汉工程大学智能机器人湖北省重点实验室,湖北武汉430073 [2]西安电子科技大学雷达信号处理国家重点实验室,陕西西安710071

出　　处：《电子学报》2012年第6期1223-1228,共6页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61172150)

摘　　要：本文提出了一种基于子空间和投影分离的三维正弦信号的频率参数估计算法.算法核心思想在于将三维采样数据阵理解为散布在某一维上的二维切片矩阵列,以此来实现对三维采样数据阵列的降维处理,并通过构造投影矩阵来实现信号中各分量的分离,算法可实现信号各分量三维参数自动配对.在整个估计过程中,算法所用矩阵都维持原采样数据阵规模,因而算法整体运算量较小.计算机仿真结果表明所提算法在性能及运算复杂度上均要优于三维情况下的IMDF算法及HOSVD算法.A computationally efficient method that combines the subspace and projection separation approaches is proposed for 3D frequency estimation of multiple sinusoids. From the perspective that 3D sinusoids is a series of 2D sliced matrices dis tributed in the third dimension, frequency parameters in the first dimension are estimated by applying the conventional MUSICbased method to the covariance of the series of slice in the third dimemion. Subsequently, a set of projection separation matrices is con structed to separate and estimate the frequency parameters in the second and the third dimension through a set of slices distributed in the third and the second dimension,respectively. The 3D frequency parameters are automatically paired.For only the original scale of sampled data array is utilized, and the proposed method is more computationally attractive. Computer simulaliom show that the proposed algorithm is superior to IMDF and HOSVD methods of 3D case in terms of estimation performance and computational complexity.

关 键 词：三维频率估计 子空间方法 投影矩阵分离方法 

分 类 号：TN911.7[电子电信—通信与信息系统]