Abstract
In this paper a fast method for blind identification of periodic sources is presented. In the well-known second order blind identification method, the information is extracted from instantaneous mixtures by simultaneously diagonalizing several time-delayed covariance matrices, however, the delays are chosen arbitrarily. This imposes computational cost which is linearly related to the number of covariance matrices. Statistical characteristics of periodic sources are exploited here to develop a method to effectively choose the appropriate delays in which the diagonalization takes place. Detail theory together with the corresponding theorems have been presented. Software simulations verify the superior performance of the algorithm in the face of different noise and frequency variation levels over alternative methods. © EURASIP, 2010.