基于LD分解的空间谱估计技术

A NEW SPATIAL SPECTRUM ESTIMATION TECHNIQUE BASED ON LD FACTORIZATION OF THE ARRAY COVARIANCE MATRIX

提出了一种基于对阵列协方差矩阵进行 LD 分解的空间谱估计算法。对一大类阵列来说,该算法对空间谱估计方法的工程实现问题具有十分重要的实际意义。文中证明了特殊Hernitian 矩阵 LD 分解的原理,给出了求分解因式的递推公式,然后,由 LD 分解找出的信号子空间,得出了估计信号源方向的空间谱计算公式。理论分析和模拟结果表明,LD 分解和原有的特征分解法相比,更易于用专用数字信号处理片实现,有更小的计算量,需更少的存储单元。

This paper presents a noval spatial spectrum estimation algorithm based on LDfactorization of the array covariance matrix.For a lot of array,this method is very important toimplement spatial spectrum estimator of DF(direction finding).This paper shows that theprinciple of special Hermitian matrix LD factorization,gives a series of recurisive formulasto obtain the LD factorization,then in accordance with extracted signal subspace,forms spatialspectrum estimation formulas for DF.Theoretical analysis and simulation results illustrate,that compared with the eigen analysis method,LD factorization method is more easier to im-plement with dedicafed digital signal processing chips,and is less in load and complexity ofcomputation,needs less store space.