I should caution that the 802.3ae committees and the IEEE's high speed study group (HSSG) have spent a lot of time going over an almost endless number of distance-wavelength-modality explorations. Any reference to fiber distances wrt GbE or 10 GbE going forward needs to be qualified as per type of utilization and its intended venue. Some even include SONET enveloping, as in the 10 Gb/s Ethernet in OC-192 variant.
10 Gb, for example, has an assortment of proposals still pending (but mostly passed by now), and some proprietary adaptations have already been unleashed by their promoters, for very short reach (VSR) applications, such as those intended for central offices, colos and network access points (NAPs) where ISPs congregate and hand off to backbone providers.
Most of these have been whittled down already, but the old rules of thumb, being limited in their absolute number, no longer apply as they once did.
Today there are variations in transport, some standardized and some still proprietary, that depend on straight SMF, MMF, DWDM, CWDM and even some that use multiple physical paths (multiple fiber strands) to contend with, when discussing distance limitations for higher order Ethernet links. FWIW. |
