BLIND SOURCE SEPARATION BASED ON TIME-FREQUENCY SIGNAL REPRESENTATIONS

Blind source separation consists of recovering a set of signals of whi ch only instantaneous linear mixtures are observed, Thus far, this pro blem has been solved using statistical information available on the so urce signals. This paper introduces a new blind source separation appr oach exploiting the difference in the time-frequency (t-f) signatures of the sources to be separated. The approach is based on the diagonali zation of a combined set of ''spatial t-f distributions.'' In contrast to existing techniques, the proposed approach allows the separation o f Gaussian sources with identical spectral shape but with different t- f localization properties, The effects of spreading the noise power wh ile localizing the source energy in the t-f domain amounts to increasi ng the robustness of the proposed approach with respect to noise and, hence, improved performance. Asymptotic performance analysis and numer ical simulations are provided.