Other noise limits that quickly swamp the quantum noise are those I mentioned at the beginning: Raman, amplified spontaneous emission from in-line optical amplifiers, Rayleigh scattering, Brillouin scattering. Again, the calculation will need to be modified for those depending on the exact link parameters.
These noises tend to increase in proportion to the link length, and therefore one often sees bandwidth-distance products quoted for "he-man" announcements of fibre capacity in experiements (e.g. 1 terrabit per second over a 100km link; the same link should roughly take 2TBps over 50km , 4tBPs over 25km and so on until the quantum noise limits everything).
As above, the bandwidth limit B also has an inverse dependence on fibre length, but a zero length fibre transmission will still be marred by the quantum noise of the link's source. So the true dependence on link length L of the capacity C, taking this into account, will be something like C=C1L0+L where L0 is something much less than a kilometer and accounts for the source's quantum noise.
Current figures quoted are tens of terrabits per fibre core - (see here - I'm sorry I don't have a better reference for this, it has been some time since telecommunication technology was wonted to me). Even higher figures can be gotten for short fibres (kilometres in length) with multimoded cores so that the power density in the core is not so constraining.
fiber optic engineer salary
The disadvantage is that dispersion (now difference between transmission speeds of the fibre's bound eigenmodes) is now the limiting factor, thus only short fibres can be used. For a single mode fibre, only chromatic dispersion, as discussed by roadrunner66, is present. This can be effectively cancelled when the link dispersion is known (as it is for long haul links) by imparting the inverse dispersion at either the sending or receiving end using a Bragg grating device.
These noises tend to increase in proportion to the link length, and therefore one often sees bandwidth-distance products quoted for "he-man" announcements of fibre capacity in experiements (e.g. 1 terrabit per second over a 100km link; the same link should roughly take 2TBps over 50km , 4tBPs over 25km and so on until the quantum noise limits everything).
As above, the bandwidth limit B also has an inverse dependence on fibre length, but a zero length fibre transmission will still be marred by the quantum noise of the link's source. So the true dependence on link length L of the capacity C, taking this into account, will be something like C=C1L0+L where L0 is something much less than a kilometer and accounts for the source's quantum noise.
Current figures quoted are tens of terrabits per fibre core - (see here - I'm sorry I don't have a better reference for this, it has been some time since telecommunication technology was wonted to me). Even higher figures can be gotten for short fibres (kilometres in length) with multimoded cores so that the power density in the core is not so constraining.
fiber optic engineer salary
The disadvantage is that dispersion (now difference between transmission speeds of the fibre's bound eigenmodes) is now the limiting factor, thus only short fibres can be used. For a single mode fibre, only chromatic dispersion, as discussed by roadrunner66, is present. This can be effectively cancelled when the link dispersion is known (as it is for long haul links) by imparting the inverse dispersion at either the sending or receiving end using a Bragg grating device.
No comments:
Post a Comment