Title :
An adaptive IIR algorithm with unimodal performance surfaces
Author :
Karni, Shlomo ; Zeng, Gengsheng
Author_Institution :
Dept. of Electr. & Comput. Eng., New Mexico Univ., Albuquerque, NM, USA
fDate :
2/1/1988 12:00:00 AM
Abstract :
The authors develop an adaptive IIR (infinite-impulse response) algorithm by means of least squares inverses; the algorithm also guarantees the resulting IIR system to be stable. This IIR algorithm consists of two subalgorithms, i.e. the all-pole QN( z) and all-zero PN(z) algorithms. Both are with quadratic performance surfaces and stable. The iterations for QN(z) and PN(z ) can be executed simultaneously or alternatively. The all-pole part is independent of PN(z), but the all-zero part depends on QN(z) at each iteration. In many cases, the all-pole model itself is a fairly good approximation of an IIR system
Keywords :
digital filters; filtering and prediction theory; IIR filter; adaptive IIR algorithm; all-pole algorithm; all-pole model; all-zone algorithm; infinite-impulse response; iterations; least squares inverses; quadratic performance surfaces; unimodal performance surfaces; Adaptive algorithm; Adaptive systems; Finite impulse response filter; Gold; Hydrogen; IIR filters; Least squares methods; Polynomials; Signal processing algorithms; Speech processing;
Journal_Title :
Acoustics, Speech and Signal Processing, IEEE Transactions on