How to get the BER from the PAM4 eye diagram in MMF system

[Yizhou Liu]

Hi,

I did a PAM4 through MMF simulation. But I cannot get the correct BER after regenerate the signal. I want to know how to get the BER from the PAM4 eye diagram. Or there is another tool can do it in optisystem.

[Damian Marek]

Hi Yizhou,

As an M-Ary signal there are two main ways. One using a BER Test Set that compares an input sequence to and output sequence exactly, but needs long bit sequences to calculate very small BERs.

Or like in the attached example, using a constellation diagram and a custom target file also attached. I made a PAM4 scheme and a custom target file to calculate the symbol error rate and the error vector magnitude. Make sure before running you correctly define the directory of the .dat file in the Constellation Visualizer properties.

Cheers

[Yizhou Liu]

Hi Damian,

Could you attach the target file again. There is no .dat file in the attachments. Also where I can see the symbol error rate and the error vector magnitude.

Thanks

[Damian Marek]

Hey,

Sorry about that! In the project file I checked the results to display underneath the electrical constellation diagram.

Regards

[Yizhou Liu]

Hi Damian,

Thank you for your reply. By the way, is it possible to use multi mode fiber in your PAM direct detection. I tried to replace the Laser and SMF in your diagram with Spatial VCSEL, Parabolic-index MMF and Spatial PIN photo-diode. The BER Test Set gives me results, but the Electrical Constellation Visualizer seems not work well. Could you help me check my setting in the attached files.

Thank you

[Yizhou Liu]

I used the same target file you provided to me last time.

[Damian Marek]

The problem here is that the Laser has been changed and is now operating at different powers for the different symbols. I had tweaked the .dat file to match the other Laser settings. You can either tweak the new laser to have the same power levels as the original (better option) or modify the .dat file to reflect the new levels.

Regards

[Santiago Echeverri]

Hello Damian,

I was wondering if you know of a way to script the Target values so that they scale proportionally to the output power at the photodiode. I am thinking about doing a sensitivity curve where the estimated BER is plotted against received power or even signal to noise ratio.

One Idea that comes to mind is to generate many .dat files with known scaling, but If you know of a more automatic way it would be very useful.

Regards,
Santiago

[Author]

Posts