Reduced complexity blind unitary prewhitening with application to blind source separation

Eigenvalue decomposition (EVD) of the sample data covariance matrix is, typically, used for calculating the whitening matrix and prewhitening the noisy signals. An important problem here is to reduce the computational complexity of the EVD of the complex-valued sample data covariance matrix. In this paper, we show that the complexity of the prewhitening step for complex-valued signals can be reduced approximately by a factor of four when the real-valued EVD is used instead of the complex-valued. Such complexity reduction can be achieved for any axis-symmetric array. For such class of arrays it enables real-time implementation of the prewhitening step for complex-valued signals. The performance of the proposed procedure is shown in application to a blind source separation (BSS) problem