Abstract:In the first part of the paper, the one-dimensional functional equation g(y(t)) = cg(z(t)) with known functions y and z and constant c is studied. Its indeterminacies are calculated, and an algorithm for approximating g is proposed. Then, this linearization identification algorithm is applied to the postnonlinear blind source separation (BSS) problem. In the case of bounded sources, a self-organizing map is used to approximate the boundary, and the postnonlinearity estimation is reduced to the one-dimensional equation from above. For supergaussian sources, the density maxima are interpolated by performing linear BSS within concentric rings. Again postnonlinearity es- timation using the ring approximation separates the mixtures. I. INTRODUCTION Linearization identification solves the functional equation 1. Intuitively, it calculates g such that the two functions y and z are scalings of each other after application of g i.e. the linear relationship between g y and g z already determines g. It can be applied for separating postnonlinear mixtures given various restrictions to the sources. We demonstrate this in the case of independent bounded sources. Other possible applications include assumptions such as sparseness or nonnegativity of the sources. This paper is organized as follows: In the next section, we calculate the indeterminacies of linearization identification and give an algorithm for estimating g. Section III describes the postnonlinear BSS problem, for which an algorithm based on linearization identification is proposed in section IV. The first step of this algorithm consists of a so-called curve detection, which is described in the case of bounded sources in section V and in the case of supergaussian sources in section VI. Computer simulations are then given in section VII.

A Linearization Approach to Blind Separation of Convolution Mixture Signals

Rapid Second-Older Blind Source Separation Algorithm for Vibration Signals

Weierstrass Approach to Blind Source Separation of Multiple Nonlinearly Mixed Signals

Algorithm for Multichannel Blind Deconvolution Based on Subspace Decomposition

Postnonlinear blind source separation via linearization identification

Blind Source Separation with Compressively Sensed Based on SBL

A smart universal single-channel blind source separation method and applications

Blind Separation Using Second Order Statistics for Non-Stationary Signals

Blind Source Separation of Nonlinearly Constrained Mixed Sources Using Polynomial Series Reversion

Approach of blind source separation based on experience character function and its application

A Solution to Blind Separation of Convolutive Communication Mixtures in Frequency Domain

Nonlinear Blind Separation Using an RBF Network Model

Neural Network Approaches to Nonlinear Blind Source Separation.

Blind Source Separation Algorithm for Communication Complex Signals in Communication Reconnaissance

An Explicit Connection Between Independent Vector Analysis and Tensor Decomposition in Blind Source Separation

A Fixed-Point Algorithm For Blind Separation Of Temporally Correlated Sources

Blind source separation by sparse decomposition in a signal dictionary

A New Mixing Matrix Identification Algorithm for Underdetermined Blind Source Separation

Particle filtering algorithms for single-channel blind separation of convolutionally coded signals

Blind Source Separation of Interfering Signals in Analog Circuits

Quasi-Blind Source Separation Algorithm for Convolutive Mixture of Speech