FIXED-POINT ERROR ANALYSIS OF THE QR-RECURSIVE LEAST-SQUARE ALGORITHM

This paper presents a detailed analysis of the QR-RLS algorithm in fin ite and infinite precision environments. The mean squared values of al l internal variables in steady state are derived. They are useful to d imension the wordlength of each variable of the algorithm. The recursi ve equations describing the quantization error propagation are derived and the stability conditions are presented. Then, we derive analytica l expressions for the mean squared values of the deviations in the int ernal variables of the algorithm, for fixed-point arithmetic. In parti cular, new analytical expressions for the excess of mean squared error and for the variance of the deviation in the tap coefficients are der ived. All the analytical results are confirmed to be accurate through computer simulations.