Mike & lml, I was just doing some catch-up, and expected there would be a different reaction to the original hypothesis that there is some bandwidth choke point at the "backdoor" (trunking side) of typical ILEC central offices. I naturally can't speak for all, but for at least one ILEC in the east, this is nonsense for all but perhaps a handful of rural, non-interesting COs of a few hundred lines each.
All voice frequency trunking (paired copper) was replaced years ago first with T1, subsequently with T1, T3, etc. muxed on multiple fiber routes out of the COs. There is typically substantial dark fiber still waiting to be used, and data rates are typically not yet muxed to the maximum of the fiber capability. Engineering decisions are make as to the best way to augment capacity -- so the ultimate capacity is not there, but it is typically a 1 - 2 month job to add a lot of capacity when its needed. This is absolutely true for any large, strategic CO in my eastern ILEC. With a couple of thousand COs, there are probably some transition cases where a next planned job adds a capacity lift, but no offices are sitting on any kind of trunk-side data capacity bottleneck.
Did I seriously misunderstand the original proposition?
oh, and lml, muxing on analog fiber/coax is usually as easy as frequency division multiplexing additional RF channels with a cheap block converter. I know I don't understand what the concern is that this is somehow difficult. Of course, this is only good out to the upper cutoff frequency of the coaxial plant and passives. For new cable placed since 1994, that Fc for cable and passives is usually 1GHz, so with 750MHz implementations, there is often some headroom -- just don't try to amplify it with the existing amps, and figure that the attenuation of the cable is quite a bit higher at these upper frequencies.
|
