Finite precision analysis of the fast QRD-RLS lattice algorithm

This work derives relations for the mean squared values of the deviations in the FQRD-Lattice algorithm. The objective is determining the level of accuracy in the error signal of this adaptive filtering algorithm. It is shown that many internal variables have the same dynamic equations of the QRD-RLS-based algorithms, and results of previous infinite precision studies can be used. Simulations showing the validity of the derived equations with different precisions are included.