Blind Extraction of Chaotic Sources from White Gaussian Noise Based on a Measure of Determinism

摘要

This work presents a new method to perform blind extraction of chaotic signals mixed with stochastic sources. The technique makes use of the features underlying the generation of chaotic sources to recover a signal that is "as deterministic as possible". The method is applied to invertible and underdertemined mixture models and illustrates the potential of incorporating such a priori information about the nature of the sources in the process of blind extraction.