Optisystem simulation/ dispersion

[Rodaina Gamal Abd Elnasser]

Hello,
I am trying to simulate an IM/DD system on optisystem. I am using a CW laser with linewidth 10 MHz at 1550 nm, electroabsorption modulator and fiber length of 20km. The electrical signal i/p to the modulator should be upconverted to any subcarrier frequency in the electrical domain using Matlab. For example I can upconvert the data to 1.25 GHz or 8.75 GHz using Matlab before sending the data to optisystem.
The problem is as the subcarrier frequency increase the dispersion increases greatly i.e. if the subcarrier frequency is 1.25 GHz or baseband the dispersion effect does not appear at all but if the subcarrier frequency is 8.75 GHz, the dispersion effect is too large which is not logical for me since this change in frequency maps to very little change in wavelength and taking into consideration that the bit rate is only 1.25 Gb/s.
If i set the laser linewidth to zero, the dispersion effect does not appear at any subcarrier frequency.
Does anyone has an explanation for that ??
Thanks in advance

[Damian Marek]

I saw your post related to this one. My guess is that dispersion has nothing to do with it. You are correct that over such a short distance and without a large bandwidth signal dispersion should be very low. However, I do believe the Low Pass Filter in the PIN Photodiode is going to cutoff frequencies greater than 1.25 GHz. Have you set that low pass filter correctly?

[Rodaina Gamal Abd Elnasser]

Thanks Damian Marek for replying
The total band i am working on is 10 GHz as the sampling frequency is 20 GSample/sec. So but the bandwidth of each user is 2.5 GHz which can be anywhere in the GHz band. So if the sub-carrier frequency is around 1 GHz there is no dispersion but if the sub-carrier frequency is around 8 GHz there is large value of dispersion.
I did not a LPF to the PD but there is a LPF after electrical down-convertion (demodulation) to the baseband and it is set correctly because if i turn off the fiber dispersion in the optisystem the BER is very small

So the problem seems to be in the dispersion.
attached are the system setup and the optical spectrum when the electrical subcarrier frequency is 8.75 GHz

Attachments:

1. 

   1-1.png
   

2. 

   2-1.png
   

[Jude Ifeanyi]

I am working on Coherent detection Optical OFDM. My main focus is to tackle the chromatic dispersion on fiber as well as residual dispersion left after equalization. I don`t know how to go about it, mostly i need the software and tutors to aid me. Thanks all.

[Mohamed A. A. Abouseif]

Jude, you can download a one-month trial version of OptiSystem to help you in modeling. go to the downloads menu and install it!

thanks
Mohamed Abouseif

[Author]

Posts