re: telco deep fiber designs, variations on the theme: FSAN, NGDLC, Hybrids
AHhaha, before answering your upstream questions, let me first state that deep fiber schemes are covered by about three or four dominant template designs, and variations thereof, as seen as the necessary differentiating criteria by vendors producing them. We see the same thing in cable TV cable modem termination system (CMTS) designs at head ends, but with many more variations in the telco realm than that of the cable TV, where only a few vendors have been able to match the CableLabs criteria.
[[Granted, the CMTS point solution is not the same as that of FTTC pedestal criteria, but they do demand certain conditions in a similar way. And T's new deeper fiber solution, where they seek to reduce the number of homes passed per node, may be taking on some proprietary attributes of their own which will need to be reconciled with the Cable Labs recommendations, at some point. I'll let others who are more knowledgeable with that area speak to that issue.]]
The primary design templates for deep fiber deployments by the ILECs are, and not necessarily in this order at this time:
- FSAN, or the Full Service Area Network scheme, which can employ SONET and ATM in the first leg out from the central office, and then employs passive optical networking (PON) fiber distribution to distributed optical networking units (ONUs) in the field...
cselt.it
- Next Generation Digital Loop Carrier, or NGDLC such as AFCI is known for (see url below).
- and then there are several other proprietary adaptations of these which combine both sets of attributes in a hybridized fashion.
Some good illustrations of NGDLC and hybrids are available in the following AFC panel, which I use here simply for illustrative purposes. Note the distinctions between FTTC, fiber in the loop (FITL), fiber in rural areas, etc.
afc.com
They all share some common characteristics, however. Primary among the LCDs are:
- the use of fiber on the longer sections to replace the long copper feeder cables between the Central Office and the targeted distribution area, a.k.a. the telco's serving area;
- Each makes use of a new intermediary set of networking elements between the fiber feeder just mentioned and the local distribution twisted pair cables to structure. These devices are variously placed in those green telco field pedestals, or in quansut huts, or in some form of shared or leased facilities (in denser areas this may be an easement in an apartment building or commercial structure such as in the FITL model in the above AFC url, for example) somewhere in the immediate proximity to customers.
- some of them make it possible to choose the last mile medium from among multiple forms of extension to the home (twisted pair, coax, and possibly wireless in the future), but this is usually relegated to twisted pair copper for the time being, between the intermediate field device (at the pedestal or field hut) and the home.
"How does fiber to the pedestal improve DSL distribution outside of absolute distance improvement?"
In various ways for some implementations, and in a more limited way for others. Any improvement would be relative to where you are, and which form of delivery was used (choose from the above).
In the extreme case, deep fiber (of the type being deployed in Atlanta, say) might facilitate 100 Mb/s IEEE protocols at some point, or GbE extension to the home, where only a maximum of 12Mb/s or at best 52Mb/s VDSL might have been possible in the best of cases using standard DSL techniques.
And if you look at the FSAN model, it's also possible to deliver full NTSC video to the home, and other more traditional forms of services through the use of very high speed SONET rings which feed ONUs, which, in turn use ATM to the home. Many of these frameworks are two to five years old, and only a few of them have actually been deployed to date.
"The number of pedestals would have to be substantially larger than the equivalent number of nodes in the enhanced version of Att cable distribution, based on your twisted pair local load carrying capacity representation.
That would depend on the medium used. For published payload capabilities such as DSL and ISDN, the copper twisted pair's balance characteristic are amazingly resilient to outside influences, and because of its design thwarts many of those ingress factors which one would ordinarily assign to interference. It's got to do with the twist effect and how intrusive magnetic fields cancel themselves, routinely. As long as the resistive/ capacitive/ inductive, or RCL properties, on each conductor of the pair are equal to one another (to within manageable tolerances), then the effects of interference are self-nullified.
"If aggregating shared lines runs into mutual interference problems over a 50 house loop, the case for residential DSL weakens on a cost basis vis-a-vis cable."
I think I just answered that question above. Although I should also add that because of the closer placement of the pedestals to the client, much higher speeds could also be achieved with even less interference than on those longer loops from the central office that existed before.
"How do CLECs provision a pedestal? How many get to share the pedestal or is this answered in the FCC ruling?"
Aahh.. these issues fall under a separate set of docket numbers, designed to address the "separate subsidiary creation by the ILEC" and the unbundled network element (UNE) issues which are still pending. These are, IMO, going to be directly impacted or will have a direct impact on, the copper sharing order at some point. Stay tuned.
Regards, Frank Coluccio |
