Browse Prior Art Database

Fast and Accurate Computation of the Amplitude of Quadrature Pairs

IP.com Disclosure Number: IPCOM000121623D
Original Publication Date: 1991-Sep-01
Included in the Prior Art Database: 2005-Apr-03
Document File: 2 page(s) / 58K

Publishing Venue

IBM

Related People

Raghavan, SA: AUTHOR

Abstract

Accurate determination of the amplitude of a quadrature pair is a necessity in many signal processing applications. It is also useful in digital PES demodulation of the servo system for a partial response maximum likelihood (PRML) system. The exact determination of a quadrature pair involves doubling the number of bits and also finding the square root, which is a sequential process and, hence, time-consuming. The new method described in this article is very attractive from the standpoint of accuracy and speed. This method is called "2-Step Least Square Estimate" (2-LSE). This estimate is obtained by using piecewise linear approximation of a function. The number of steps in the estimate refer to the number of linear functions used to approximate the given function.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 66% of the total text.

Fast and Accurate Computation of the Amplitude of Quadrature Pairs

      Accurate determination of the amplitude of a quadrature
pair is a necessity in many signal processing applications.  It is
also useful in digital PES demodulation of the servo system for a
partial response maximum likelihood (PRML) system.  The exact
determination of a quadrature pair involves doubling the number of
bits and also finding the square root, which is a sequential process
and, hence, time-consuming.  The new method described in this article
is very attractive from the standpoint of accuracy and speed. This
method is called "2-Step Least Square Estimate" (2-LSE).  This
estimate is obtained by using piecewise linear approximation of a
function.  The number of steps in the estimate refer to the number of
linear functions used to approximate the given function.  For a given
number of steps, the best linear approximation is obtained by
minimizing the total mean squared error of the estimate and, hence,
the name LSE.  This estimate has a worst- case error of only 1.32%

      The 2-LSE for the amplitude of quadrature pairs is shown
graphically in Fig. 1.  If we call the quadrature pair X and Y, and
their amplitude A, then A is given by

      Let us assume that X and Y are non-negative (if they are
negative, their absolute values can be taken).  If we call A' the
2-LSE of A, then A' can be written in the form
      A' = C1X+C2Y
in either of the two pieces of the linear a...