改进的频域盲分离排序不确定性消除算法  

Improved Algorithm to Eliminate Permutation Indetermination of Frequency Domain Blind Source Separation

作　　者：王卫华[1] 黄凤岗[1] 王桐[2] 

机构地区：[1]哈尔滨工程大学计算机科学与技术学院,哈尔滨150001 [2]哈尔滨工程大学信息与通信工程学院,哈尔滨150001

出　　处：《系统仿真学报》2009年第2期496-499,共4页Journal of System Simulation

摘　　要：讨论了频域盲源分离算法中的排序不确定性问题。根据相邻频点上同一个信号的频谱幅度相关大于不同信号频谱幅度相关的特点,提出了一个改进的基于相邻频点幅度相关性的排序不确定性消除算法。在两个相邻频点上构造一个幅度相关矩阵,再根据相关性大小将幅度相关矩阵转换为转置矩阵。该方法可以实时处理卷积信号频域盲源分离算法的排序不确定问题,在对每个频点进行独立分量分析的同时确定各分量的顺序。该算法效率高、鲁棒性好。仿真实验表明了算法的有效性。The permutation problem in the frequency domain blind source separation was discussed. By utilizing the characteristic that amplitude correlation between neighbor bins of the same signal is better than different signals, an improved method based on the amplitude correlation between neighbor bins to eliminate the permutation indetermination was proposed. An amplitude correlation matrix between two neighbor bins was constructed, and it was transformed into a permutation matrix according to the value of amplitude correlation. This method can deal with the permutation problem of convolutive mixture blind separation in real time. It determines permutation of independent components （ICs） as soon as the independent component analysis algorithm is implemented at a frequency bin. This method is robust and fast. Simulation experiments demonstrate a good performance.

关 键 词：频域盲源分离 卷积混合 排序不确定性 幅度相关 

分 类 号：TN912.3[电子电信—通信与信息系统]