\code{insitu meas <ch\_1> <ch\_2> <samples> <measurements>}& starts a
delayCoefficient ($\alpha$) measurement.
Measurements alternate between ch\_1 and ch\_2.
For each channel round trip delay samples are
recorded. DelayCoefficient ($\alpha$) can be
determined for each ch\_1, ch\_2 measurement
set.
Available if \texttt{CONFIG\_INSITU\_CALIB} is
set. \\
\code{insitu stop}& stops automatic sweep over number of channels\\
\code{ip get}&\\
...
...
@@ -2854,7 +2863,7 @@ It adds the following commands:
\begin{lstlisting}
sfp tune <channel|wavelength>
insitu <on|off|crtt|run|stop>
insitu <on|off|crtt|run|meas|stop>
ptp sync_int <n>
ptp delay_req_int <n>
ptrack insitu
...
...
@@ -2879,7 +2888,7 @@ Current wavelength: 1563.85 nm
This subsection describes obligatory and optional steps to get input for further
calculations.
As a first step it is recommended to disable the assymetry of a link. It can be
As a first step it is recommended to disable the asymmetry of a link. It can be
done by setting alpha to 0 for the used SFP in the local database.
To achieve this, get the Vendor PN of the SFP by:
\begin{lstlisting}
...
...
@@ -2893,7 +2902,7 @@ Then add it to the SFPs database:
wrc# sfp add FTLX6872MCC 20000030000000
\end{lstlisting}
Please note that in this example deltas are set to an example values. Last two
zeros set the alpha to 0(disable assymetry).
zeros set the alpha to 0(disable asymmetry).
The second optional step is to avoid White Rabbit synchronization, only layer-1
syntonization will be used.
...
...
@@ -3062,6 +3071,110 @@ To generate a plot out of the generated files the following scripts can be used:
\end{figure}
\FloatBarrier
\subsection{Measure delayCoefficient ($\alpha$)}
The delayCoefficient ($\alpha$) can be measured using the 'three one-way delays (TOWD) procedure'.
Round trip delay measurements are sampled for different laser wavelengths (tunable SFP channels).
Depending on the location of the tunable laser SFP (at the timeTransmitter or at the timeReceiver) the reverse or forward channel is kept at the same (common) wavelength.
First channel 1 is selected and a number of round trip delay samples are recorded.
Next channel 2 is selected and round trip delay samples are recorded.
Multiple of these channel 1, channel 2 measurement sets can be recorded for statistics.
During the measurements, no asymmetry must be active and there must be only layer-1 syntonization:
\begin{lstlisting}
sfp add FTLX6872MCC 20000030000000
ptrack insitu
\end{lstlisting}
Please note that in this example deltas are set to an example values. Last two
zeros set the alpha to 0(disable asymmetry).
Now alternate between channel 1 and 2 and sample the round trip delays:
\item\texttt{ch\_1} is the channel used for wavelength 1
\item\texttt{ch\_2} is the channel used for wavelength 2
\item\texttt{samples} defines number of round trip delay samples to collect
\item\texttt{measurements} defines number channel 1, 2 measurement sets to be collected
\end{itemize*}
The following command alternates channels 4 and channel 53 and measures 5 round trip delay samples for each.
This is repeated until 10 sets of measurements for channel 4 and channel 53 are recorded:
\begin{lstlisting}
insitu meas 453510
\end{lstlisting}
The measurement can generate a significant amount of data to the console.
The output of the virtual console can be redirected to a file by e.g. command:
\begin{lstlisting}
$ ./spec-vuart 2>&1 | tee output_file.txt
\end{lstlisting}
The intitu meas ouput file can be analyzed using a python script.
This script analyzes the recorded round trip delays and calculates the delayCoefficient from the sets of averaged round trip delay measurements for ch\_1 and ch\_2.
The laser wavelengths in use are needed in order to be able to calculate the delayCoefficient:
