Browse Prior Art Database

Power Estimator for Automatic Gain Control

IP.com Disclosure Number: IPCOM000106802D
Original Publication Date: 1993-Dec-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 2 page(s) / 79K

Publishing Venue

IBM

Related People

Davis, GT: AUTHOR [+4]

Abstract

Disclosed is a method for sampling the power value of a signal before initializing an Automatic Gain Control (AGC) circuit.

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

Power Estimator for Automatic Gain Control

      Disclosed is a method for sampling the power value of a signal
before initializing an Automatic Gain Control (AGC) circuit.

      The AGC circuit allows the processing of a signal over a range
of power values by scaling the signal as its power level increases or
decreases.  After a valid signal power level is detected by a power
estimator, AGC parameters corresponding to this power level are
initialized for immediate convergence.  The power estimator uses the
output of a Hilbert filter, yielding both real and imaginary
components for accurate calculation of the signal power.  This method
is used in a Digital Signal Processing (DSP) application of voice
grade signals, allowing AGC sampling to be done at a low rate
requiring relatively little processing power.  An example of such an
application is found in a modem algorithm.

      A power estimator of this kind ensures that the power is
stable, so that noise glitches are ignored.  If the power changes by
a certain percentage, which provides a variable threshold, the scan
rate is reset to wait for the stabilization of power at a new level.

      After a signal processing function is enabled in a computing
system, the power estimator is enabled.  The signal is fed through a
Hilbert filter, which is a bandpass filter letting through only those
frequencies which are of interest to be used for power calculation.
The Hilbert filter returns a first signal corresponding to the real
component of the input signal, and a second signal corresponding to
the imaginary component of the input signal.  For example, if the
input signal is of the form Asin(wt), the real signal is of the form
Asin(wt), while the imaginary signal is of the form Acos(wt).  In the
power estimating process, these signals are each squared, with the
results being added to yield, in this example of a sinusoidal signal,
the square of the amplitude, A.  After the power is calculated in
this way, the AGC gain value is determined using, for example, a
logarithmic table lookup procedure.

      After the code is first initialized, the FPWRGSCN() state, in
which the power is monitored, is set.  When the power rises above a
lower threshold level and becomes stable, (i.e., it does not vary by
mor...