Infinite precision analysis of the QR-recursive least squares algorithm

A detailed analysis is presented of the QR-decomposition/recursive-least-square (RLS) algorithm in infinite precision, when the input signal samples are not correlated. Equations for the dynamic range of the algorithm are derived. A program in C is written in order to simulate the QR-RLS algorithm implemented with fixed-point arithmetic in order to verify the accuracy of the formulas presented. In all examples, the experimental set up consists of the system identification application where both the input signal and measurement noise are pseudorandom sequences with normal distribution and zero-mean.<
>