- This topic has 12 replies, 6 voices, and was last updated 9 years, 8 months ago by Ravil.
-
AuthorPosts
-
-
April 25, 2014 at 6:41 am #11029Nitin K LoharParticipant
Hi everyone,
Can I implement a protocol (say, GPON or EPON) using Optiwave and do the simulations? Can I define the frame structure and packet sizes in Optiwave? -
April 29, 2014 at 12:09 pm #11231Damian MarekParticipant
Hi,
OptiSystem can simulate passive optical network performance. I attached a picture of a GPON layout. Unfortunately, OptiSystem is more of a physical layer simulation and so you cannot directly define frame structure and packet sizes.
-
January 20, 2015 at 7:32 am #17262MarioParticipant
Hello Damian,
I am trying to make a simple simulation of a GPON and i found your GPON-Layout in previous post. I have made and found everything least the spliter. Can you please show me how to create it? Or do you have any layout done that you can send me to take a look on them?
Many thanks in advance
Ps: another thing, how to rotate the elements(like flip vertically)?
Best Regards
Mário Dias -
January 20, 2015 at 7:32 am #17264MarioParticipant
Hello,
For the previous post by me i put a 1xN Splitter Bidirectionalthat i think is the appropriate one.
Now after i have done a very simple GPON layout(Not Raman amplification) when i run the project and when i check the BER analyzers it gaves a strange graph(i think is not normal). Do i have something wrong in the project?
Also in the upstream BER analyzer i cant see any data.
Project in attach
Thanks in advance
Best Regards
Mário Dias-
January 20, 2015 at 7:56 am #17270MarioParticipant
Hello,
I think i had the wrong layout parameters(bit rate, etc). Now i think i am on the way, At least ONU BER looks normal.
But i still dont know what is happening with upstream in OLT, no data arrives there. do you know what can be worng?
(in attach the current layout)
Thanks in advance
Best Regards
Mário Dias -
January 20, 2015 at 9:56 am #17273Damian MarekParticipant
Hi,
For bidirectional simulations you usually need to increase the simulation iterations, I changed yours to 5 and got results. I also removed the Buffer Selector (you can switch iterations in the signal visualizer).
One problem though is that you have two users sending data upstream on the same carrier wavelength, so their signals are getting mixed up.
Regards,
Damian -
January 20, 2015 at 1:33 pm #17277MarioParticipant
Hello Damian,
Many thanks for your fast answer. Just one question, is not normal the users use the same wavelength to send data upstream?is not the standard? shouldnt ODelay prevent the signal mixing?
Thanks
Best Regards
Mário Dias -
January 21, 2015 at 9:15 am #17282Damian MarekParticipant
If you want to run a time division multiplexing simulation, you will have to run it a little differently. The Optical Delay used here is more of a simulation tool. Since all components need an input before they can simulate, you need to run the simulation for multiple iterations and for the first iteration the delay component will emit a null signal.
This tutorial probably explains it better than I can in a quick reply:
Lesson 5: Bidirectional Simulation — Working with Multiple Iterations
Regards
-
-
-
March 2, 2015 at 7:32 am #17933Shahbaz QureshiParticipant
Hi Damian,
The picture you have attached for GPON shows 3R regenarator, is it okay to use it? Wouldn’t it over come the actual results? I guess it is not used practically?-
March 2, 2015 at 10:05 am #18027Damian MarekParticipant
Yes the purpose of the 3R Regenerator is to simplify the connections in a design. Normally you would need the original bit sequence and generated electrical signal along with the output transmitted signal to generate a realistic eye diagram. This component removes the need to make multiple forks and connections from the input. However, sometimes it can fail for very noisy systems or in some other cases where the global bit rate does not match the received bit rate. In these cases I generally use the input signals for the Eye Diagram Visualizer.
Regards
-
-
March 3, 2015 at 12:10 am #18119RavilParticipant
Hi Damian, I have a question about usage of 3R regenerator as a splitter and Bit Sequence Generator. Based on many schemes I’ve seen, it is the most common way for bit error detection by means of eye diagram. Am I right?
Can you, please, explain how accurate (and different from using fork and connection to a real generator) will be the result obtained in such scheme?-
March 10, 2015 at 5:34 am #18470Parikhit DuttaParticipant
Hello Ravil,
I had a similar doubt on using a 3R Generator + BER analyzer over connecting the data signal from the transmitter end to the BER analyzer at the receiver. From Damien’s post I am assuming that we use a 3R generator mostly for simplification of the layout.
To compare a 3R Generator + BER analyzer with a BER analyzer with connections from the transmitter, I used them simultaneously in a sample project simulating a simple RoF System. What I observed is that the Min. BER using a 3R Generator was much higher than that without a 3R Generator, e.g., 10^-5 as against 10^-8 respectively.
I also compared the two schemes for a back-to-back connection against a system with a 50 km long SMF between the transmitter and receiver. The Min. BER using 3R Generator was always higher. Damien has also pointed out that the noise may play a part. May be that may have a role to play because of using MZMs, power combiners, EDFAs in the system I designed. It is attached here.
-
-
March 10, 2015 at 12:17 pm #18485RavilParticipant
Hi Parikhit!
Thank you very much for your feedback! The results make sense to me, and I’ll definitely take them into consideration for my future design. Since I’m traveling right now, I can’t look at your scheme. However, I think that we should test these schemes only on back to back topology. In other case, channel noise and, especially, signal distortion will have an influence on BER.
-
-
AuthorPosts
- You must be logged in to reply to this topic.