Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Optimal Infinite Impulse Response Filter Design by Mean Field Annealing

IP.com Disclosure Number: IPCOM000109532D
Original Publication Date: 1992-Sep-01
Included in the Prior Art Database: 2005-Mar-24
Document File: 3 page(s) / 126K

Publishing Venue

IBM

Related People

Nobakht, RA: AUTHOR

Abstract

This article describes a new technique for designing stable Infinite Impulse Response (IIR) filters. An arbitrary order IIR filter can be designed which will very closely model the magnitude and the phase response of a Finite Impulse Response (FIR) filter. Higher performance improvements can be obtained with the use of the proposed technique. Also, since the designed IIR filter will require much less number of coefficients necessary to achieve the same level of performance for a given FIR filter, many hardware elements or software instructions could be saved.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Optimal Infinite Impulse Response Filter Design by Mean Field Annealing

       This article describes a new technique for designing
stable Infinite Impulse Response (IIR) filters.  An arbitrary order
IIR filter can be designed which will very closely model the
magnitude and the phase response of a Finite Impulse Response (FIR)
filter.  Higher performance improvements can be obtained with the use
of the proposed technique.  Also, since the designed IIR filter will
require much less number of coefficients necessary to achieve the
same level of performance for a given FIR filter, many hardware
elements or software instructions could be saved.

      From a set of fixed FIR filter coefficients an IIR filter of
arbitrary order N is designed.  The order of the filter, N, depends
on the complexity of the filter and the choice of the designer.  The
figure shows the basic diagram of the design optimization process.
The FIR filter realization follows directly from the convolution sum
relationship written in the form

                            (Image Omitted)

                                                       (1)
where the desired signal y(n) is generated by convolving a
pseudo-random sequence, x(n), by the weight vector w* with elements
w*(i).  The IIR filter model can be described as the following
z-transform representation
                                                       (2)
where H(z) is the z-transform of h(n) and the prediction signal mu(n)
can be described as
                                                       (3)

      Finally, the prediction error is calculated as
                                                       (4)
and the root mean square (RMS) of this error is calculated on the Nth
error sample alone or on several of the previous p...