基于GRD-Hough变换的多高斯源盲分离算法

盲源分离有一个重要假设:源信号最多只含一个高斯信号。否则,基于统计量的盲分离算法性能会恶化。本文从广义矩形分布出发,通过把时域中的一维信号映射到二维的时-频表示来提供信号的频谱内容随时间变化的信息,并对时频谱进行Hough变换处理,利用不同高斯源的时频分布差异性,避开统计量提出了一种能分离多个高斯源的盲分离算法,扩展了盲源分离的应用领域。

Blind separation algorithm based on GRD-Hough transform for multi-Gaussian sources

There is an essential assumption in the blind source separation(BSS) that the sources do not allow more than one Gaussian signal.Otherwise,the separation algorithms based on statistics would worse the performance of the model.This paper introduces a new BSS approach exploiting the difference in the time–frequency(t-f) signatures of the sources to be separated.The approach is based on the remodeled generalized rectangular distribution to obtain t-f signatures,and then localized the t-f spectrum by Hough transform.The proposed approach allows the separation of Gaussian sources with identical spectral shape but with different t-f localization properties,and hence extends the application field of BSS.