摘要
盲信号分离是近几年才发展起来,用于解决从混合观测数据中分离源信号的一门新技术,已在许多领域获得了广泛应用。本文介绍了盲分离的主要理论和两大类实现方法——独立分量分析和非线性主分量分析,并在此基础上介绍了实现盲信号分离的不同算法、在非线性混合情况下的算法以及盲信号分离将来的发展方向。
Blind source separation (BSS) is a recently developed methodology used to separate unknown source signals from their mixtures. It has been applied to many fields widely and effectively. The theory and two types of implementation methods-independent component analysis (ICA) and nonlinear principal component analysis (nonlinear PCA) are introduced. Then the methods of blind source separation in the condition of linear mixing and nonlinear mixing are discussed. Finally the future of blind source separation research is prospected.
出处
《信息与电子工程》
2003年第1期69-79,共11页
information and electronic engineering
基金
国家自然科学基金资助项目(10276005)
关键词
育信号分离
独立分量分析
最小互信息法
非线性主分量分析
最大熵法
blind source separation
independent component analysis
nonlinear principle component analysis
minimum mutual information
maximum entropy