Narrow band source separation in wide band context applications to array signal processing

Abstract

Blind source separation is now a well known problem. Various methods have been proposed for instantaneous and convolutive mixtures of sources. Conventional antenna array processing techniques are based on the use of second order statistics but rest on restrictive assumptions. Thus, when a priori informations about the propagation or the geometry of the array are not available, the model can be generalized to a blind sources separation model. It supposes the statistical independence of the sources and their non-gaussianity. In this paper, we focus on the narrow band source separation problem embedded in wide band jammers. We show that the JADE algorithm made for instantaneous mixture is still valid in a wide band context where only the signals of interest are narrow-band. We also prove that a wide band signal tends to occupy all the degrees of freedom of the covariance matrix and modifies the signal subspace dimension.