OSNR vs Log(BER)

[Dr. Dhiman Kakati]

Dear all, I need to find the OSNR vs Log(BER) plot so how can I set OSNR as my wish and find the corresponding log(BER) for the same.
please help.

Regards

[Alessandro Festa]

Dear Dhiman, you need to be a little more specific…for which particular case do you want to simulate BER vs OSNR? Can you attach a design?

[Dr. Dhiman Kakati]

I am not attaching the design file of the system that I have but I can give you a similar design image. this is approximately same as mine. I need your suggestion.

Regards

[alistu]

Hi Dhiman,

You can use an optical power meter at the receiver, where the optical power is going to be transformed into electrical power. This gives you power of signal and noise separately. But the set OSNR component can set the OSNR by adding a noise to the signal according to your defined OSNR value.

Regards

[Dr. Dhiman Kakati]

Dear Alistu, I am not getting the first line you have mentioned “You can use an optical power meter at the receiver, where the optical power is going to be transformed into electrical power. This gives you power of signal and noise separately.” if it is possible please add same snapshot.

Thank you

Regards

[alistu]

By the above sentence, I simply meant to say that the optical power meter should be used at the beginning of the receiver side. One of the first things done at the receiver is to transform light into electricity. Then the power will be electrical and you will not be able to achieve Optical-SNR anymore. If you want, I will attach some snapshot but it is fairly easy: Just put it after the line and before the receiver.

Regards

[Dr. Dhiman Kakati]

I think for calculating OSNR we can use the “set OSNR” component (let 25dB) but when i connect the log(BER) or BER becomes arounfd 0.45. that is not acceptable (so high). So I need help.

Regards

[alistu]

As far as I know, the OSNR Set creates the OSNR you define by adding a noise floor to the signal. Therefore, I think if the signal is not strong enough or if the received OSNR is almost less than the amount you have set for it to be, there is nothing this component can do to adjust the SNR. I am not sure about it, though.

[Dr. Dhiman Kakati]

Oh, Thanks Alistu. I basically need to plot the OSNR vs log(BER). but due to high BER I am unable get a good curve plot.

Thanks

[Roman Ibragimov]

Oh. its very interesting question. I want to know too< how we can calculate BER, if I will have values OSNR. Of course, it will be theoretical curve, however I will have possibility to match BER vs OSNR from Optisystem with theoretical limit

[alistu]

Hi Roman,

I think your question is a bit different from that of Dhiman’s. In order to know how to theoretically calculate BER with respect to the SNR, I suggest that you go through the paper I am attaching to my post here. It mostly talks about Q-factor, but then you can calculate BER with regard to the Q-factor.

Regards

[Dr. Dhiman Kakati]

Dear Roman,
You can calculate the BER in many ways one way is to use the BER test set component that is available in the component library.
the input terminal of BER test set connected to the decoder output or if if you are using a dual polarization scheme them connect to the output of the parallel to serial converter.

you can also use WDM analyzer to check the OSNR.
Given the OSNR, the empirical formula to calculate BER for single fiber is

Log10 (BER) = 10.7-1.45 (OSNR)

Example:
Assume that OSNR = 14.5 dB
ThenLog10 (BER) = 10.7-1.45 (14.5) = -10.30
Therefore BER = 10^(-10.30)
BER is approx 10^-10

I think it helps

Regards

[Roman Ibragimov]

Dear Dhiman. Its very empirical formula, because in this case I can’t consider modulation format (DP-QPSK, SP-QPSK etc).
Do you have source, where i can read about formula?

[Dr. Dhiman Kakati]

I am attaching the pdf here.

Regards

[Dr. Dhiman Kakati]

regards

[Alessandro Festa]

Hi Dhiman, the document does not specify for which bitrate this formula is valid, and we know that BER vs OSNR does depend on bitrate!

[Dr. Dhiman Kakati]

Oh I am not sure about that so we have to wait for others to reply for the same.

Regards

[gaganpreet Kaur]

i just want to confirm i have attached a dual port analyzer and getting my OSNR as high as 93 . my firt concern are these high values of OSNR realistic. secondly when i attach BER analyzer at receiver at high bit rate of >100gbps no results are shown. in that case will this empirical formula give me correct BER for OSNR achieved.

[Alessandro Festa]

Hi, 93dB OSNR seems quite unrealistic…for which design did you get it?

[Dr. Dhiman Kakati]

I am attached the OSNR versus Log(BER) curve. Please go thrugh and please comment how to find this.
Thanks

[Roman Ibragimov]

Do you have description for this figure. How many channels? It’s single channel or DWDM system. which is modulation format (NRZ, DPSK, QPSK, 16QAM, OFDM+QPSK)? How did you get this figure? from experiment?

[Dr. Dhiman Kakati]

you can go through the pdf, the sanpshot is from inside the pdf.

Regards

[Alessandro Festa]

Hi Dhiman, without you posting the .osd design file it is quite difficult to help you on this system…

[Dr. Dhiman Kakati]

only one channel, No WDM or multiplexing.

Regards

[Dr. Dhiman Kakati]

I am attached the OSNR versus Log(BER) curve. Please go through and please comment how to find this.
Thanks

[Author]

Posts