Eye diagram problem

[UMAR FAROOQ]

Hello members i used DQPSK modulation format in 4 channel wdm system design.Its eye diagram is not fine at DQPSK receiver side how shall i improve it.Image is attached.

Attachments:

1. 

   Untitled1.png
   

[Alessandro Festa]

Hi Umar, can you attach the design file or give an explanation of the design itself?

Regards,

Alessandro

[Dr. Dhiman Kakati]

Dear Umar, Please attache herewith your design file and reference paper pdf so that I can check, I have also worked before in this topic.

Regards

[UMAR FAROOQ]

Actually i have to calculate the power budget of my whole system.I have 4 channel wdm system on transmitter side i applied DQPSK modulation format after multiplexing then i have 50 km SMF link then WDM DEMUX after that i use power splitter one signal goes towards DQPSK receiver and the other to RZ OOK Transmitter for remodulation then these 4 remodulated signals are multiplexed and then SMF then DEMUX and then to OOK receiver and BER analyzer.I use DQPSK in downstream and OOK in upstream for the calculation of my power budget and this eye diagram is on DQPSK receiver side.
Please suggest which modulation format is good for long range PON and what are the improvements possible in this sytem.
Regards
Umar Farooq

[alistu]

Hi Umar,

Since I don’t have access to your simulation file, I can only make some general suggestions that I have formerly made in other topics. You can use DCF to reduce the dispersion effect. As for the launched power, you can increase it to the point where the nonlinearities kick in. You can use optical amplifier after the transmission line, but not more than the fiber attenuation.

Some advanced modulation formats such as OFDM have shown better performance especially against dispersion and can be considered to be used in PONs (especially in downstream, with upstream being OOK for example and using remodulation techniques), but since “cost” is an important factor and should be taken into consideration, it is not scientific to say which modulation format is “better”. Otherwise, all the researchers would have focused on one or two modulation formats.

Regards

[UMAR FAROOQ]

I use DCF but it made no change same problem exists.

[alistu]

What about my other suggestions regarding how to use optical amplifiers and fiber launched power? It would be better if you upload your simulation file here (if it is possible for you of course) so that we have access to the implementation and be able to make the changes to make it work OK.

Regards

[UMAR FAROOQ]

According to your suggestions i use optical amplifier and change the input power but my eye diagram remains same.Upstream eye diagram is fine but downstream eye diagram remains same when i use DQPSK at transmitter.When i use DPSK its all fine but reach is limited.

[alistu]

I am assuming you cannot upload the file due to some reasons, so can you tell me the linewidth of the lasers, gain of the optical amplifiers and your launched power? Also can you tell me how remodulation has been done? (I hope you realize it is hard to troubleshoot a system without having access to it though)

[UMAR FAROOQ]

Line width 0.2 MHz,Gain of amplifier 20 db,launched power 0 dbm. At ONU, a 3dB optical splitter divides downstream into two parts. An
intensity modulation technique of 10-Gbps OOK is used to re-modulate the first half of the power splitter to generate upstream data signal while the second half goes to DQPSK OR DPSK receiver.This re-modulation takes place by the received downstream signal which is either DPSK or DQPSK. I know its hard to understand but currently i am working on this project and it is not possible for me at this time to upload simulation.
just see my eye diagram and tell why one eye coincide with another eye as i have clear eye opening on up streaming side.

[alistu]

The eye diagram consisting of two eyes corresponds to the coded multilevel signal. So you would just have to improve the Q-factor and BER. If you are using 20dB amplifier, I assume you have a 100 km transmission line. You can further increase launched power until the system reaches nonlinear effects (that’s usually when simulation takes a long time). You can decrease the linewidth to, say, 0.1MHz (unless you are reproducing the work of a paper). You can also consider using DSP component to compensate for nonlinearities, dispersion and some other unwanted effects.

Regards

[UMAR FAROOQ]

Thanks for your suggestions i will make changes and when when i got my desired results i will share in forum.

[alistu]

You’re welcome. I hope they will be helpful.

Regards

[Author]

Posts