Based on ICA (Independent Component Analylsis) principle

a fast leaning algorithm for blind signal separation is presented.By seeking the local extrema of the fourth-order cumulants (i.e.kurtosis coefficients) of a linear combination of the observed variables

the model and the process of this algorithm are obtained and then used for the experiment of blind signal separation.The results of the experiment show that this algorithm has a great deal advantages in blind signal separation and features extraction such as fast convergence and needless in any dynamic parameter and the like.The algorithm can separates all independent components from blind source signals

which is non-Gaussian distribution.The algorithm is a new

highly efficient and reliable method in signal processing.