- This topic has 7 replies, 2 voices, and was last updated 9 years ago by
Alessandro Festa.
-
AuthorPosts
-
-
November 2, 2015 at 9:09 am #27349
alistuParticipantHi all,
I have implemented a WDM system using OptiSystem. When I increase the bit rate (and therefore I have to increase the channel spacing), the channels that are further from the optical fiber reference wavelength have worse constellations than the channels near the fiber reference wavelength.
I have attached a screenshot from the WDM spectrum after the fiber link as you can see. I have also attached an image showing constellation diagrams of an eight-channel WDM (channels one to eight, from left to right), while the fiber reference frequency is the same as the central frequency of simulated band. As you can see, the signal is more corrupted as the channel as farther from the central frequency of the fiber, even though I have used a DCF in the line.
I would like to know why this signal degradation is happening while I have used a fiber as DCF with the same characteristics and when the dispersion becomes noticeable in the main fiber, dispersion in the second fiber with the same (yet negative) slope and dispersion value should compensate for it. Am I right?
Thank you all in advance.
Regards,
Alistu -
November 3, 2015 at 2:13 am #27381
Alessandro Festa
ParticipantHi Alitsu, your project file seems to be corrupted, when I open it I can only see the orpopagation fiber, but not the input WDM signals…can you posti it again?
I hope I can help you.Alessandro
-
November 3, 2015 at 2:30 am #27382alistuParticipant
Hi Alessandro,
Unfortunately, I cannot share the whole project file. I have only sent the fiber link so as to let you know of fiber characteristics. I have used a WDM-OFDM system, but it is not really important, since the problem has something to do with fiber dispersion and must be there in any WDM system with the specified channel spacing that uses the same fiber link. Thank you for your consideration.
Regards
-
-
November 3, 2015 at 3:05 am #27383Participant
Hi Alitsu, ok I understand your point.
For WDM cases, I usually do not use Constant Dispersion parameters, but I use Disp vs wavelength from one of the files built-in in Optiwave.
You can try to use one file for the propagation fiber dispersion, and create the one for the DCF by simply multiplying each dispersion number * (-5). This should improve your simulation.Let me know if this works.
Alessandro
-
November 3, 2015 at 3:27 am #27384alistuParticipant
Thank you very much for the suggestion. However, what is the problem with the current link? As you can see in the SMF used (which is the OptiSystem default SMF with some minor changes), the dispersion is not constant for all channels and the channels symbol rotations are different. Besides, even for a channel with constant dispersion, the DCF with my dispersion settings should work, as it is like multiplying numbers by -5 like to your suggestion and if it is not, should I change the settings in any way?
-
-
November 3, 2015 at 3:39 am #27385Participant
I am not sure your way with constant Dispersion actually works well.
You are using 16ps/nm and 0.08slope, and for DCF -80 and -0.45 slope. Suppose you have one signal 1nm away from the reference: in 5km propagation its dispersion is 80.4, while its compensation is -80.45. I am afraid that at high bitrate the 0.05 difference may affect the performances…-
November 3, 2015 at 7:26 am #27397alistuParticipant
Thank you Alessandro. There was this simple miscalculation that had occurred only in this simulation file and had confused me. If the amount of dispersion parameters are exactly set by multiplying the values of the main fiber by -5, there would be no problem and it would have the exact same effect as what you have suggested about manipulating file values in your former comments.
Best regards
-
November 3, 2015 at 7:55 am #27398Participant
You are welcome Alitsu. I encountered a similar issue for 100G WDM systems before, so now I make sure to exactly compensate all the dispersion for all wavelength at high bitrates.
Regards,
Alessandro
-
-
-
-
AuthorPosts
- You must be logged in to reply to this topic.