Atec Boonton-4400A-4500A - Manual (Page 4)

CCDF of 8 carriers spaced 700 kHz
apart from 1.8 GHz to 1.8049 GHz
and phase aligned for highest peak
power.
PDF of 8 carriers spaced 700 kHz
apart from 1.8 GHz to 1.88049 GHz
and phase aligned for highest peak
power.
CUMULATIVE DISTRIBUTION
FUNCTION (CDF)
The Model 4500A CDF is a plot of
power on the vertical axis versus
probability on the horizontal axis. The
logarithmic power capability allows the
entire dynamic range to be viewed at one
time or a small region to be expanded up
to 0.1 dB/division for detailed analysis.
Likewise, the probability scale can be
expanded up to 0.1%/division. Both scales
can be offset over their full range to bring
any portion of the magnified CDF into
view. For every possible power level (p)
within the total dynamic range, the CDF
shows the probability in percent that the
measured power is less than or equal to p.
The maximum power level during the
entire run will appear at 100% probability.
The adjustable markers allow the power
to be read at any probability value along
the curve, interpolating to a resolution of
0.01%. Adjustable reference lines allow
the probability to be read for any power
level to a resolution of 0.01 dBm.
Summary results are continually updated
and appear at the top of the display, as
follows:
· Peak Power
· Average Power
· Peak-to-Average Ratio
· Total Elapsed Time
· Total Number of Samples
· Tolerance
· Reference Line Value
Alternatively, a TEXT display may be
selected to display the above in addition to
the following:
· Minimum Power
· Dynamic Range
· Confidence
· Marker Position and Value
· Reference Line Position and Value
When the second channel is installed,
the CDFs for the two channels can be
compared using the Math channel. CDFs
can also be saved into REF1 and REF2
and to the floppy disk. The confidence
level of the CDF is determined by the
number of samples taken. For many
purposes, useful results are obtained in
just a few seconds of run time. For events
with a low probability of occurrence, the
run time must be increased to ensure
capture.
COMPLEMENTARY
CUMULATIVE DISTRIBUTION
FUNCTION
The same histogram data can be used
to display the complement of the CDF or
CCDF, which is also called the 1-CDF
because of the manner in which it is
calculated. For every possible power level
(p) within the totally dynamic range, the
CCDF shows the probability in percent that
the measured power is greater than p.
This definition flips the CDF curve from left
to right placing the maximum power value
on the zero percentage axis. This form is
often more convenient to use, but contains
the same information as the CDF.
PROBABILITY DENSITY
FUNCTION (PDF)
The cumulative distribution
function can be further
processed to obtain an
approximate probability density
function or PDF. The PDF
shows the relative occurrence
of the various power levels.
The probability that the power
lies between any two values on
the power axis is equal to the
integral of the PDF between
the two values. When the
difference between the two
values is equal to zero, the
integral of the PDF is also zero.
This means that it is not
possible to determine the
probability of occurrence of a
particular power level. Strictly
speaking, a discrete array such
as the histogram of power
values described above does not have a
density function. However, it is useful for
qualitative analysis to create an
approximate PDF from the first-order
derivative of the CDF. The process of
differentiation exaggerates any errors in
the CDF values. For this reason,
measurements should always be made
from the CDF or CCDF. The PDF is useful
for visualizing QAM and other signals that
have a finite number of discrete levels. For
random noise signals, A Gaussian
amplitude distribution will appear as a
Rayleigh distribution in the linear power
domain.
The TEXT display of measurements for
the PDF is the same as for the CDF and
CCDF. Adjustable markers and reference
lines are also provided; there are no
marker measurements in the PDF mode.
BOONTON ELECTRONICS (A Wireless Telecom Group Company)
P.O. Box 465, Parsippany, New Jersey 07054-0465 · Tel: (973) 386-9696 · Fax: (973) 386-9191 · E-mail: [email protected] · Website: www.boonton.com