作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2012, Vol. 38 ›› Issue (06): 187-189. doi: 10.3969/j.issn.1000-3428.2012.06.061

• 人工智能及识别技术 • 上一篇    下一篇

基于K-SCA假设的欠定盲源分离

杨 文 1,张宏怡 2,普杰信 1   

  1. (1. 河南科技大学电子信息工程学院,河南 洛阳 471000;2. 厦门理工学院电子与电气工程系,福建 厦门 361024)
  • 收稿日期:2011-05-25 出版日期:2012-03-20 发布日期:2012-03-20
  • 作者简介:杨 文(1986-),男,硕士研究生,主研方向:盲信号处理,模式识别;张宏怡、普杰信,教授、博士

Underdetermined Blind Source Separation Based on K-SCA Assumption

YANG Wen 1, ZHANG Hong-yi 2, PU Jie-xin 1   

  1. (1. Electronics & Information Engineering College, Henan University of Science and Technology, Luoyang 471000, China; 2. Department of Electronic and Electric Engineering, Xiamen University of Technology, Xiamen 361024, China)
  • Received:2011-05-25 Online:2012-03-20 Published:2012-03-20

摘要: 传统稀疏算法对信号的稀疏程度要求高、抗噪能力差。针对该问题,从K-SCA假设出发,提出一种基于超平面隶属度函数的欠定盲源分离算法。该函数基于局部统计,具有良好的抗噪性能,适用于噪声和信号稀疏程度较低条件下的信号分离。实验结果表明,相比同类算法,该算法对信号稀疏要求低、分离精度高、容噪能力强。

关键词: K-SCA假设, 欠定盲源分离, 超平面聚类, 稀疏分析

Abstract: To solve the problems of the traditional algorithm based on sparse analysis, the good sparsity requirement of signals, unsatisfactory results in the noise case, the method based on hyperplane membership function is proposed to estimate the mixing matrix fast and precisely according to the assumption of K-SCA. For the good locality of the function, the noise problem which the classical method can not solve can be solved well using the proposed method. Experimental result shows that compared with other methods, the required key condition on sparsity of the sources can be considerably relaxed.

Key words: K-SCA assumption, Underdetermined Blind Source Separation(BSS), hyperplane clustering, sparse analysis

中图分类号: