论文部分内容阅读
盲信号分离是20世纪最后10年迅速发展起来的研究领域,其目的是仅利用观测信号估计或分离出源信号。首先介绍了盲信号分离的基本概念、发展历史,接着对具有线性瞬时混合、线性卷积混合和非线性混合三种混合方式的盲信号分离模型进行了阐述,进一步总结了目前针对正定、欠定、和单通道混合信号的盲分离方法,对热点方法的原理和特点进行了概括,最后展望了盲信号分离技术发展的难点、趋势和应用方向。
Blind signal separation is a rapidly developing area of research in the last decade of the 20th century. Its purpose is to estimate or separate only the source signal using only observed signals. Firstly, the basic concepts of blind signal separation and its history are introduced. Then, a blind signal separation model with three kinds of hybrid methods, ie, linear instantaneous mixture, linear convolutional mixture and non-linear mixture, is introduced. At the same time, , And the single-channel mixed signal blind separation method, summarizes the principle and characteristics of the hot-spot method, and finally presents the difficulties, trends and application directions of the development of blind signal separation technology.