An Algorithm of Blind Signal Separation Based on Kurtosis
Yingmin Wang · Modern Electronic Technique · 2005
The Blind Source Separation (BSS) problem consists of recovery sources from the observed signals without adequate a prior knowledge.Independent Component Analysis(ICA)is a statistical method for transforming an observed multidimensional random vector into components that are statistically as independent from each other as possible, and it is one of the most important approaches to the BSS problem. An algorithm for BSS based on kurtosis is presented in this paper, which is simpler and easier to implement in solving Givens Matrix compared with Comon′s. As no other assumption of the Probability Density Function (PDF) is made, the algorithm can be used for virtually any PDF. It is also extended to the general BSS problems in light of Comon′s pair wise principle. The simulation justifies its effectiveness.