基于循环优化的矩阵联合对角化算法及在盲源分离中的应用

被引:4
作者
李昌利
曹嘉毅
机构
[1] 广东海洋大学信息学院
关键词
盲源分离; 联合对角化; 非对角元素; 对角元素; 循环优化;
D O I
10.15961/j.jsuese.2011.05.024
中图分类号
TN911.7 [信号处理];
学科分类号
0711 ; 080401 ; 080402 ;
摘要
通过对一组与盲源分离中观测信号有关的目标矩阵进行近似联合对角化,可以有效解决盲源分离问题。提出一个非正交的循环优化的联合对角化算法。待优化的2个函数分别为矩阵对角化处理后的非对角元素的平方和与对角元素的平方和,通过对两者的循环最小化和最大化处理即可获得对角化器。对比仿真实验表明本文算法可以获得更好的盲源分离效果。
引用
收藏
页码:159 / 162
页数:4
相关论文
共 12 条
[1]  
A comparative study of approximatejoint diagonalization algorithms for blind source separation inpresence of additive noise. Degerine S,,Kane E. IEEE Transactions on Signal Processing . 2007
[2]  
Nonorthogonal joint diagonalization freeof degenerate solution. Li X L,Zhang X D. IEEE Transactions on Signal Processing . 2007
[3]  
A fast fixed-point algorithm for independent component analysis. Hyvarinen A,Oja E. Neural Computation . 1997
[4]   非正交联合对角化盲源分离算法 [J].
张华 ;
冯大政 ;
聂卫科 ;
徐先峰 .
西安电子科技大学学报, 2008, (01) :27-31+70
[5]  
Fast approximate joint diagonaliza-tion of positive definite Hermitian matrices. Todros K,Tabrikian J. Proceed-ings of IEEE International Conference on Acoustics,Speech,and Signal Processing . 2007
[6]  
On using exact joint diagonalization for noniterativeapproximate joint diagonalization. Yeredor A. IEEE Signal Processing Letters . 2005
[7]  
Non-orthogonal joint diagonalization/ zero-diagonalization for source separation based on time-frequency distributions. Fadaili E M,Moreau T N,Moreau E. IEEE Transactions on Signal Processing . 2007
[8]  
Non-orthogonal joint diagonalization in the least-squares sense with application in blind source sepa- ration. Yeredor A. IEEE Transactions on Signal Processing . 2002
[9]  
Adaptive Blind Signal and Image Processing: Learning Algorithms and Applications. Cichocki A,Amari S. . 2002
[10]  
A blind source separation technique using second-order statistics. Belouchrani A,Abed-Meraim K,Cardoso JF,et al. IEEE Transactions on Signal Processing . 1997