Goef, thanks for the clarification and for some good additional observations. I didn't think that it was your intent to discount NRZ in the previous message, I was just seeking your clarification in order to make sure I wasn't missing something.
----
Somewhere in between straightforward NRZ and soliton coding there lie several intermediate, or midway, remedies for extending distance-speed reach. One remedy is the use of pre- and post- emphasis in optical pulse shaping, in order to compensate for the nonlinearities (the phase shift characteristics, and the derivatives of same) of the medium.
When using "pre-emphasis" there is the intentional distortion of the pulse shape introduced by the transmitting end point (or in post emphasis at the receiving point, or both), which means that when the pulse is "launched" it resembles the end-to-end distortion characteristic of the medium itself, only equal and opposite to it.
In this fashion, when the pulse comes out at the other end of the pipe it has been compensated, already, by the medium's own distortion profile. This is suitable up to a point (distance-wise), rendering the NRZ signal states highly discernable for detection purposes and decoding at the optical receiving element.
The second method has been the constant efforts on the parts of the glass makers themselves, improving dispersion/shifting characteristics of their fiber so that we now have a wide variety of options to select from when choosing a fiber profile to match the throughput and distance requirements --among other considerations-- for a particular application design.
Neither of these entirely eliminates the properties which affect the overall reach limitations of a strand, or the need at some point to use optical amplification or electronic regeneration (or "o-e-o," for those who prefer to spell the whole process out) at some point, on extremely long routes.
Instead, these measures simply lengthen distances to new norms. And again, these assume that NRZ is the coding scheme of choice.
I have to wonder how the linear improvements in fiber products, which were designed around the needs of current coding schemes (such as NRZ), will fair when soliton transport comes to be. Anyone have any ideas about this? I would welcome more comments here by yourself, and by those who might be more familiar with the progress taking place in soliton research.
TIA, Frank |
