On The Performance Comparison of Gradient Type Joint-Process Est

It is well known that a joint process estimator consists of two struc- tures, namely the orthogonalizer and the regression filter. In literature, orthogonalization step is performed either by or- thogonal transformations or by linear predictors.

While the orthogonal transformations do not preserve entropy; the predictors, such as the lattice, do preserve it. However, the steady-state performance of such linear predictors is not as good as those of the orthogonal transformations. Lattice filters do not perform perfect orthogonalization when they operate as gradient-based adaptive predictors.