Effective Algorithms for Regressor Based Adaptive Infinite Impulse Response Filtering

IMPULSE RESPONSE FILTERING Emrah Acar Dept. of Electrical and Computer Engineering Carnegie Mellon University Pittsburgh, PA 15213 [email protected] Orhan Ar kan Dept. of Electrical Engineering Bilkent University Ankara, Turkey 06533 [email protected] ABSTRACT To take advantage of fast converging multi{channel recursive least squares algorithms, we propose an adaptive IIR system structure consisting of two parts: a two{channel FIR adaptive lter whose parameters are updated by rotation{ based multi{channel least squares lattice (QR{MLSL) algorithm, and an adaptive regressor which provides more reliable estimates to the original system output based on previous values of the adaptive system output and noisy observation of the original system output. Two di erent regressors are investigated and robust ways of adaptation of the regressor parameters are proposed. Based on extensive set of simulations, it is shown that the proposed algorithms converge faster to more reliable parameter estimates than LMS type algorithms. 1. THE REGRESSOR BASED IIR ADAPTIVE FILTER STRUCTURE As shown in Fig. 1, in a typical adaptive ltering application, input, x(n), and noisy output, d(n), of an unknown system are available for processing by an adaptive system to provide estimates, y(n), to the output of the unknown system as time progresses. We assume that the unknown +