Bias Removal Algorithm for 2-D Equation Error AdaptiveIIR Filters

  • Authors:

  • Maha Shadaydeh;Masayuki Kawamata

  • Affiliations:

  • Department of Electronic Engineering, Graduate School of Engineering, Tohoku University, Sendai, 980-8579, Japan;Department of Electronic Engineering, Graduate School of Engineering, Tohoku University, Sendai, 980-8579, Japan

  • Venue:
  • Multidimensional Systems and Signal Processing
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

Thispaper proposes a bias removal algorithm for equation error-based2-D adaptive cascade IIR filters with separable denominator function.As well known, equation error-based adaptive IIR filtering algorithmshave the advantages of fast convergence and unimodal mean-square-errorsurface. These advantages, however, come along with the drawbackof biased parameter estimates in the presence of measurementnoise. The adaptive filter structure in the proposed algorithmis based on the concept of backpropagating the desired signalthrough a cascade of the denominator vertical and horizontalsections. To handle the bias problem, the proposed algorithmuses a scaled value of the output error of each of the cascadedsections as an estimate for the measurement noise embedded inthe signal part of the coefficient-update procedure of that section.Thus, while maintaining the advantages of easy stability monitoring,fast convergence, and low computational load, the effect of themeasurement noise is suppressed. Input-Output stability analysisis carried out, and the constraints required to maintain stabilityare derived. Simulation examples are presented to support theeffectiveness and the usability of the proposed bias removalalgorithm in 2-D system identification and image enhancementapplications.