DQPSK WDM

[Urvi pathak]

Hi,
I am working on DQPSK WDM system and not getting proprer output.Here i am attaching a osd file please guide me to get proper output.

[Urvi pathak]

Here i am attaching my osd file

[Dr. Dhiman Kakati]

Hi Urvi Pathak,
I think you have made the design using optisystem 14 so I am unable to open it, Can you please provide me with the snapshot of he system, I can surely help you because I have also implemented DQPSK in one of my design. If your system is large in size please attach snapshot of transmitter channel and receiver separately. Thanking you.

Regards,
Dhiman

[Ranjeet Kumar]

Hi Urvi pathak,
You can use symmetrical compensation technique instead of post or pre compensation technique for better results.
As in this design you have use post compensation technique, so i suggest you to use symmetrical compensation technique…SMF+DCF+SMF.
It gives good ber and q-factor for long haul communication and for high data rate too.

[Urvi pathak]

Hi.
Thanks rather and ranjit will definitely try ur suggested changes and hope it will help I will let u know as I got any result thanks.

[Urvi pathak]

Hi dhiman
Here I m sending snap shot of my transmitter, optical link and receiver. I m working on 4 channel wdm each with 10gbps and compensating dispersion with dcf. Channel spacing 50GHz. 193.414thz to 193.564thz.

[Urvi pathak]

Here I m sending transmitter design.

[Urvi pathak]

Receiver design and optical link.

[Dr. Dhiman Kakati]

Hello Urvi Pathak,
Thank you for providing me with the attachment. I will try to resolve your issue. I need some more time.

Regards,
Dhiman

[Dr. Dhiman Kakati]

can anyone tell me when this kind of eye diagram comes as shown in the attached image, what may be the possible reason for this. Thanking you.

Regards,
Dhiman

[Urvi pathak]

Thanks dhiman
Waiting for your response..

[Dr. Dhiman Kakati]

Hi Urvi pathak,
I am also not getting the expected output. rather I am getting an eye diagram as attached in the reply no. 33931. Still looking into the problem and once I get the solution I will upload the design file. without using WDM the system is giving very good result but when I implement the same transmitter and receiver in WDM the eye diagram is as in abouve component. Thanks for the patience.

Regards,
Dhiman

[Ranjeet Kumar]

Hi Urvi,
Go through the links which may help you to get more idea about dqpsk-wdm system and which help you in your research works.
http://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=5588472
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=5758591&newsearch=true&queryText=dqpsk-wdm
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=1283616&newsearch=true&queryText=dqpsk-wdm
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=4800951&newsearch=true&queryText=dqpsk-wdm

Thanks and regards
Ranjeet

[Dr. Dhiman Kakati]

Hello Ranjeet Kumar,
If you provide with the papers pdf filed attached here this will be more beneficial to me as I dont have IEEE access ate present. Thanks in advance.

Regards,
Dhiman

[Sanjeev kumar]

Hello Urvi mam,
Could you please elaborate your problem by attching the snapshot of output response as i have also working on it and i think optimum bit rate setting for layout window is very much important consideration to get the good output response.

[Urvi pathak]

Hi dhiman ,
I am also getting perfect result in case of one channel with 40Gbps but when implemented in WDM it is not working. not getting problem.Thanks for trying and hopefully you find the problem with design…

[Dr. Dhiman Kakati]

Hello Urvi Pathak,

In my case the problem is with the demodulator, up to the demodulator everything is ok and as expected (Spectrum and signal power), but after the demodulator the constellation is also not coming (just few signal points are seen). I don’t thing so that there should be problem with the demodulator because without WDM everything is working fine. The problem may be at the WDM multiplexer and demultiplexer in the filter bandwidth, or at the transmitter laser frequency spacing. try changing these parameters.

Regards,
Dhiman

[Urvi pathak]

Hi Sanjeev,

As you told I have set optimum bitrate of layout to 40Gbps..In output I am getting Q factor of hardly 2 and getting any of one output either I or Q channel. When I monitor constellation diagram then it is not proper all phases I got in one quadrant only.

[Urvi pathak]

Hi ranjeet,

I tried your suggestion of symmetric compensation but it is also not working.

[Sanjeev kumar]

As i am in office right now. but i would like to ask what parameters u have set for the layout window. try setting sequence length 1024 and samples per bit 64. Also reduce the bit rate for each channel to 10Gbps. How many channels are there in your designed mux? try solving with the given parameters.
Thanks
Sanjeev

[kevin xiang]

You can try to use our PM combiner,or PM splitter, PM WDM in your system. we produce 405nm,532nm,635nm,650nm~1650nm splitter,WDM, such as 532/635nm PMWDM etc, which are used in OCT, QPSK ,EDFA, etc.

kevin
eamail; kevin.xiang@comcore.com
skype: kevin.hsiang

[Dr. Dhiman Kakati]

Hi Kevin Xiang,
I am not getting your point, Will you please elaborate. Why you have suggested polarization mode splitter, How it is going to help (except for twice the bit rate) we have not implemented polarization splitting technique here.

” we produce 405nm,532nm,635nm,650nm~1650nm splitter,WDM, such as 532/635nm PMWDM etc, which are used in OCT, QPSK ,EDFA, etc.” are you referring to practical implementation or practical design not a simulation platform..? looking forward for your kind response. Thanking you.

Regards,
Dhiman

[Urvi pathak]

Hi sanjeev,

Here i am sending you snap shot of my output,Constellation diagram and lay out parameters.I hope you will get the problem.

[Sanjeev kumar]

Try reducing bit rate from 40Gbps to 40/No of channels. As i have seen in single channel DQPSK if we increase bit rate above 40Gbps response is not good.

[Dr. Dhiman Kakati]

I agree with Sanjeev Kumar that in my case also a single transmitter and receiver (i.e. without WDM) I am able to get a maximum Bit rate of 10 Gbps. Urvi You also should try to reduce the bit rate at each transmitter to a maximum of 10 Gbps.

Regards,
Dhiman

[Urvi pathak]

Hi all,
I am getting proper output now. Thank you for helping.. Combined solution of all worked well thank a lot.

[Dr. Dhiman Kakati]

Hello Everyone please refer to the reply by Alan Robinson of this question on research gate. may get some clue about DQPSK. I am also trying to follow what he suggested.

https://www.researchgate.net/post/Eye_Diagram_not_proper#56fce080217e2013ec58db0f

if unable to see from researchgate please let me know.

Thanking you.
Regards,
Dhiman

[Sanjeev kumar]

You are always welcome urvi mam.

[SAHIL SINGH]

HI Dhiman Kakati,

Please look into the link which you posted… I am unable to open that link…

thanks & regards

Sahil Singh

[Dr. Dhiman Kakati]

Hello Sahil Singh,
for viewing the link you need to have a research gate account. I am attaching herewith the words what I was suggested.

According to him,

1. in conventional DQPSK scheme, a delay of 1 symbol is required for direct detection1..

2. sample the waveform at 2.56 THz, with 256 samples in each of 128 symbols for a sequence length of 12.8 ns.  The symbol rate is 10 GBaud,
with 2 symbols/bit for an aggregate bit rate of 20 Gbit/s.

3. the symbol rate 5 GBaud with 64 symbols each of 200 ps period? In this case your demodulator MZ delay of 2/Bit rate is correct.

4. A 2.5 GHz pulse train would explain the zeros in the eye.

So these are the main points arose by Alan.

Regards,
Dhiman

[SAHIL SINGH]

Hi Dhiman Kakati,

Well indeed i got the acees to your link by making an account in the research gate..I was not able to access earlier as i did not have an account… Thanks for providing such valuable information and highlighting the main points… Your efforts are indeed appreciated…

Regards

Sahil Singh

[Dr. Dhiman Kakati]

You are most welcome Sahil,

Now I have managed to get a better output, my problem was mainly due to some common mistakes in layout parameters and bandwidth of the demultiplexer. I have attached here with the eye diagram and constellation, please have a look.
Alan Robinson’s suggestion was very much helpful in this case. anyways Thanks to all for discussing and look deep into Urvi’s problem. This is now gets solved.

Regards,
Dhiman

[Author]

Posts