Adaptive IIR filtering with combined regressor and combined error

Abstract

The adaptive IIR filtering is known to be the more efficient method for system identification compared with the adaptive FIR filtering but is not widely used because of the bias and stability problems of the conventional adaptive IIR filtering algorithms. This paper presents a new algorithm called the CRCE (combined regressor and combined error) algorithm that can overcome the problems of the conventional algorithms by using the combined form of the regression vector and the estimation error. By controlling the composition of the combined regressor and the combined error, the CRCE algorithm continuously adjusts the coefficient update equation to achieve convergence stability and estimation accuracy. The computer simulation results also demonstrate that the performance of the proposed algorithm is better than those of the conventional algorithms. (C) 1997 Elsevier Science B.V.