Robust Simple Hyperstable Adaptive Recursive Algorithms

O. Sezer and M. Ferdjallah (USA)

Keywords

Identification, hyperstability, LMS, pole-zero placement, SHARF, and SPR

Abstract

Hyperstability theory when simplified and adapted for digital signal processing offers a new class of IIR filters, simple hyperstable adaptive recursive filters (SHARF), which is directly related to strictly positive real (SPR) transfer functions. In this paper, a constrained class of SHARF is investigated, where SPR transfer functions are designed without any priori knowledge of the filter parameters. The method proposed associated each pole with a zero on the unit circle by pole-zero placement on the unit circle method. As an illustration, the method will be applied to a pure four poles autoregressive (AR) process using least-mean square (LMS) method.

Important Links:



Go Back