基于小波变换的ICA盲源分离算法  

ICA Blind Source Separation Algorithm Based on Wavelet Transform

作　　者：洪丹枫[1] 凌青[2] 彭娇娇 

机构地区：[1]青岛大学信息工程学院 [2]海军航空工程学院电子信息工程系 [3]92326部队

出　　处：《海军航空工程学院学报》2013年第3期237-240,共4页Journal of Naval Aeronautical and Astronautical University

摘　　要：针对ICA算法不能直接应用于单通道盲分离的问题,提出了基于小波变换的ICA盲源分离算法。该算法结合小波变换对基于信息论的ICA算法进行了改进并应用在单通道盲分离中。仿真结果给出了分离后信号的波形,经检验其相似程度,分离后信号的波形与源信号的波形较为接近。According to ICA algorithm cannot be applied in single channel blind separation directly, the single channel blind separation algorithm based on wavetel transform was presented. The algorithm combines the wave- let transform to improve ICA algorithm based on information theory, applied in single channel blind separation and could reduce the complexity of the algorithm. Finally. the outcome of simulation showed that the waveform after separation was closed to the source singnal.

关 键 词：小波变换 ICA算法 相似程度 

分 类 号：TN911.7[电子电信—通信与信息系统]