Decorrelated algorithms for faster adaptation

A new decorrelation filter is introduced on both input and desired signal to bring down:the correlation of input signal and additive noise. The new adaptive algorithms are derived and discussed in different cases. The introduced auxiliary adaptive whitening filter reduces the eigenvalue spread of the input autocorrelation matrix, thus accelerate the adaptation of the main filter. At the same time, the auxiliary filter reduces the effect of colored noise power spectra and hence reduces the misadjustment of the algorithm. The improvement obtained is considerable in most cases at the expense of a minor computational complexity.