Abstract

Many complex signals obtained from high-dimensional laser systems are not purely simple chaotic signals. There may be contaminated by different components of individual chaotic signals or quantum noise signals. The extraction of each chaotic component from mixed chaotic signals is a very attractive and important research area for the analysis of electro-magnetic signals in the brain activity, physiological signals from human body, turbulence in fluid dynamics, as well as communications applications of code division multiple access (CDMA) using chaos [1]. Recently blind source separation by using independent component analysis (ICA) has been rapidly developed and applied for many complex signals [2]. ICA is a statistical estimation technique to separate mixed complex signals based on non-Gaussianity of the probability density function (PDF) of signals.

© 2007 IEEE