Coherent OFDM

[Fady]

Why the BER equals 0 in all coherent OFDM examples in the dvanced modulation Folder !
Thanks

[Ravil]

Are you using the scheme from Optiwave website?

[Jaffar Emad Kadum]

OFDM system can support very long distances over optical link up to 2500 km, try to increase the length of the optical link in these examples and the BER will be increased as well.

Best

[Ravil]

Agree with Jaffar. Try to examine dependence of BER on distance and transmission speed increase.

[Ravil]

Please, let me know about your progress!

[Damian Marek]

The BER visualizers are in “Measured” operating mode. This means that it directly compares an input signal to the transmitted signal and counts the errors. The sequence lengths are set to 16384 bits, which makes it hard to estimate BER values smaller than 1/16384. You can try increasing the sequence length, which will increase the simulation time or change the lengths of the fibers to introduce more distortion.

[Fady]

Thanks for your replies. I have tried both increasing the sequence length and increasing the length but without success. Increasing the length will give MIN BER of 0 and reducing the length will give a MIN BER of 1.

[Damian Marek]

What length are you changing the fiber to? I get the expected min BER of 0 for short lengths.

Attachments:

1. 

   minBERProof.png
   

[Fady]

Thanks for your reply. I have made length sweep from 50 to 100 km, I got min BER of 0 for lengths from 50 km to around 75km then I got minBER of 1 for longer lengths.

[Damian Marek]

This is not entirely unexpected as the longer fiber lengths leads to more loss and noise. However, the sharp drop off of the BER from 0 to 1 suggests that at one point the OFDM demodulator is completely wrong at detecting the input. This suggests to me that the delay of the optical signal is large enough (from the length), so that the demodulator is misinterpreting the symbols of the constellation diagram, because of this phase shift. This sensitivity is a problem with the OFDM demodulator and one the OptiSystem team is fixing in the next update coming later this year.

You might be able to work through this by changing the delay compensation parameter in the OFDM Demodulator manually.

Regards

[Fady]

Thanks for your response but what are typical values should I try for the delay compensation parameter in the OFDM Demodulator?

[Ravil]

I don’t think BER should change as gradually. By the way, Fady, what is the bit rate in a channel of the system? Did you keep it the same for both scenario?

[Fady]

I kept it as in the example 10Gbps

[Ravil]

For this bit rate both distances should work fine in terms of BER with OFDM from theoretical point of view. Did you try to tune delay compensation parameter?

[Fady]

I tried several values but without success

[Jaffar Emad Kadum]

Dear Fady

Attach the example file that you you are working with and I will try to modify it

[Jaffar Emad Kadum]

Optical OFDM is very sensitive to all simulation parameters, here I attached the variation of BER with number of loops in my OFDM system which are varied from 30 to 42 loops, each loop consist from 80 km SMF and 16 km DCF with their optical amplifiers.

Attachments:

1. 

   Image.png
   

[Fady]

Thanks Jaffar, I am working on the example of coherent OFDM in the advanced modulation folder in optisystems samples.

[Fady]

Could you please post any example where you get good BER results as all BER resulsts of the examples in the samples folder are not useful.
Thank you

[Fady]

Could you please post any example where you get good BER results as all BER results of the examples in the samples folder are not useful.
Thank you

[Jaffar Emad Kadum]

Hi Fady

I attached you an UWB-OFDM WDM system and it is working very good. you can remove the UWB channel and work with OFDM system

Attachments:

1. OFDM-UWB1-WDM.osd

[Fady]

Thanks Jaffar, but the Min BER in your system is 0.496 with or without UWB.

[Damian Marek]

Hi Fady,

Using Jaffar’s project I get a low Ber of 1e-18. Which version of OptiSystem are you using?

Regards

[Jaffar Emad Kadum]

Dear Fadi
I got the results above from this file which is built using OpitSystem 13.0. Check your simulation I dont know where is the error

[Fady]

Hi Damian,
The results were obtained using optisystem 13. Thanks

[Damian Marek]

Have you changed any of the layout parameters? It still doesn’t work?

[Fady]

I have not changed any parameter, it is still not working. Thanks

[Damian Marek]

Hi Fady,

Could you take a screenshot of the simulation with the BER result circled?

Regards

[Fady]

Thanks Damian, Please find the attached screenshot of the BER result.

Attachments:

1. 

   screenshot.tif
   

[Damian Marek]

Can you attach it as a jpeg or png please?

[Fady]

Please find the attached jpeg screenshot

Attachments:

1. 

   UWB.jpg
   

[Damian Marek]

Ah ok, I was looking at the UWB receiver. In this case I would set the BER Analyzer to use the Measured algorithm, because the eye diagram won’t give you an accurate BER estimation. The reason being the binary sequence has been regenerated at the output, since you can’t really view the eye diagram of an OFDM signal.

If you look at the eye diagram you will notice that it is a “perfect” eye diagram with no deviation and no noise. In this case the BER estimation will not reflect reality.

Cheers

[Author]

Posts