Forum on optical communication systems?

Question

Computer networks

Colleagues, hello everyone.
Prompt, who where discusses the questions devoted to optical data transmission? Whether it's building, designing or maintaining xWDM systems, atmospheric lasers, PONs. Issues of holography, optical transmitters - receivers, transmission media, etc.
I would be glad for any information
PS it is strange that there is no such hub "optical systems".

Answer 1

nag.ru

Answer 2

We know, but there are still few issues connected with optics. And there are a lot of snobs :)

Answer 3

Exactly where the legs grow from.
Today I am reading an article (PDF) publications.lib.chalmers.se/records/fulltext/local_156957.pdf
And in the course of reading I understand that, alas, I don’t know what causes the signal spectrum to be cut off when passing through the ROADM cascade. This applies not only to superchannels, but also to ordinary signals, this actually became relevant after the appearance of 100G channels in the 50GHz network. I try to google and alas, nothing. In the books, to be honest, I also do not remember that this moment was covered. It remains only to ask, but the question is where :)

Answer 4

The transfer function is just a term for converting an input to an output. It is used in management theory, respectively, you can read about it in university textbooks on management theory. Or on Wikipedia :)
In general, when a signal passes through a ROADM, its spectrum changes in a very interesting way. I observed this in cascades of two MUX / DMUX 100GHz, when a “shoulder” is obtained in the spectrum from amplified noise that falls into the OADM filter band.

This is a picture from an EXFO article: www.exfo.com/Documents/TechDocuments/Application_Notes/EXFO_anote171_In-Band-OSNR-ROADM-Challenge_en.pdf
As I understand it, the problem with OSNR after ROADM has two aspects: firstly, OSNR in different channels can have a different level due to different optical signal paths in different channels; and secondly, the shape of the spectrum changes after ROADM, and to measure OSNR, this change in the shape of the spectrum must be taken into account.
In my case (in my experiments), the measured OSNR in the channel after ROADM numerically coincided with the calculated one with sufficient accuracy, but the shape of the group signal spectrum really had the described step.
Since OSNR noticeably degrades on all damped elements, it is logical to expect that an increase in the number of ROADM stages will lead to OSNR degradation.

Answer 5

Thanks, I saw this article, by the way I like the JDSU article more in terms of pictures www.teracomm.com/wp-content/themes/tctheme/technical-library-docs/fiberoptic-test/osnrroadm_wp_lab_tm_ae.pdf , although the EXFO measurement method seems more logical.
The fact that the step appears due to noise - yes. The reason why they appear, by the way, is not very clear to me (i.e. why the excessive width of the ROADM filters allows noise to pass through), let's say this is a design feature, an inherent technology defect.
In total, it turns out that for each optical channel, the bandwidth of the ROADM filter is slightly larger than the useful signal (let's say random deviation). As a result, during the passage of the cascade, the total noise on these shoulders will be somewhat greater than during the passage of one cascade. Where then does the assumption come from that the resulting width of the spectrum of the useful signal narrows? After all, it should be the other way around: signal + noise :) The spectrum should be wider, and “noisier”

Answer 6

The fact that the filters pass somewhat wider than the useful signal allows you to save the useful signal when the transmitter frequency deviates (with a slight deviation, its signal will still pass through the filter). Of course, ASE within this filter bandwidth will also pass.
And I think you can understand the resulting width of the spectrum of the useful signal by looking at the picture: the spectrum of a pure signal has the shape of a bell, the bottom of which is based on the level of spontaneous noise.
The spectrum of the signal passed through the ROADM is the same as the original bell, but it is clipped by the amplified noise that has passed through the filter. Since the shape of the original signal remains unchanged, but it is cut off at a higher noise level, it turns out that the spectrum of the useful signal has decreased.