User guide
•
The R
2
values are excellent (greater than 0.95).
• The mean levels and standard deviations returned by the Q-factor calculations
differ from the measured values.
The QBER vs. Threshold graph may look as shown below:
If you increase the
BER Threshold to include more points, you will find:
• The mean levels and standard deviations returned by the Q-factor calculations
approach the measured values.
• The Q-factor decreases.
•
The R
2
values deteriorate.
The latter is caused by increasing nonlinearity. This can be inspected in the QBER
vs. Threshold graph.
N O T E
The Q-factor calculations are disabled and flagged as "not applicable" if the R
2
value
falls below 0.75.
The R
2
value can also be seen as an indicator of how well the noise distribution
fits to Gaussian shape. It will not fit, for example, if the received signal is dominated
by cross-talk or modal noise.
6 Advanced Analysis
286 Agilent J-BERT N4903B High-Performance Serial BERT