This is going to be one of my long and typically boring ramblings in reply to your suggesting that the topic of solitons is irrelevant, or somehow obtrusive, to the mission of the Last Mile Thread.
So, I invite you and others here who are wary of such impertinence to either read on right here for a larger dose of impertinence, or to commence with a continuation of this discussion on solitons over in the FCTF, upon which I would be duly honored...
OR, in Gilder's space, where some good discussion on matters other than slow response times and subscription anomalies would be welcome at this point, even to George hiself, I'm sure...
OR, create a new thread for solitons at this time, for that matter, in which case there could be no point of contention as to its pertinence.
Whichever venue you choose, you can rest assured that you will have my deepest interest and following, as always, in due course.
But only after I've completed this tome, in deference to myself and to the other individuals here who have expressed an interest in discussing solitons in the Last Mile.
"...this particular thread, The Last Mile, is not the forum for discussions of soliton solutions or how it works in a qualitative sense as the soliton regime is geared for long haul transport--cross-country and undersea. It is much too expensive and complex for the "Last Mile""
Perhaps at this time you are correct. In its early stages of development I offer you no argument that it will be used first in the long haul. As DWDM was only justified in the long haul too, at first. Today, as you know, DWDM is not only being considered for the last mile in metro desgins by CLECs and private enterprises alike, but in intramural campus and in-building applications, as well.
Allow me to offer a few more analogies which suggest that the initial intentions behind technological exploration often leads to places which were unimagined, when their first prospective uses were being mulled:
------
During the mid- to late- Sixties, text books were describing the "T" carrier system in its still nascent form, along with its inherent DS1 format, as a short-haul remedy for "pair gain" purposes, due to earlier copper exhausts. That was, in any event, the primary motivation behind its development and eventual deployment in metropolitan carrier access and edge networks, particularly between local and tandem central offices.
At that time, there were more robust forms of "analog" transmission still being mulled, using coaxial and microwave approaches, and even the WT4 waveguide system which I alluded to in my last post (although the WT4 would also support digital hierarchies, eventually, if it would have been left alone to survive).
And let's not forget satellites, which were seen as a salvation by broadcasters and international message traffic carriers to points overseas that would otherwise have been shut off by submarine cable's bandwidth limitations, at the time.
Today, while it is still principally used for those purposes, satellites are now being viewed as a potential means towards last mile liberation in rural, and even near-urban areas, more so than at any time before. I was especially intrigued by Valueman's posts here recently concerning these capabilities, especially those using high numbers of spot beams per bird.
Getting back to T1, today T1 is all pervasive and serves as the fundamental building block that eventually led to the unavoidable, it seems, logical form factors of SONET interfacing. Now, SONET interfaces which inherit many of T1's attributes also extend to the I/Os of DWDMs and terabit switches/routers which will eventually support traffic that will be carried by, guess what? Solitons.
In 1986, digital cross connects were deemed to be the sole province of only the largest common carriers, purely on the merits of price performance metrics. Then, super T1/T3 multiplexors began to take shape by startup box makers with many of the same attributes of the add-drops found in DCSs. These muxes rapidly found their way throughout enterprise networks.
During the same time frame, T3s were thought to be required only by the largest enterprises outside of the carrier industry, and only for "last mile" aggregation purposes, to boot. Today, transoceanic T3s and their larger sisters (OC-3s and OC-12s) are bought and sold in spot markets, and through other forms of auctioning on the Internet, as well. I could carry these reincarnations and permutations of fate much further, but I think I've made my point.
Who's to say that soliton technologies wont find their way into the last mile, too, as I predict they will at some point, at least in some form? Or, that solitons wont find their way into the distribution networks which are built by the Last Mile Cable Companies and xLECs, even if only as integral parts of larger meshes?
And in the last 2000 feet, who is to say that solitons wont be used to exploit cheaper materials [whose properties today are simply too lossy and "non-linear" for traditional line coding schemes], than today's silica based fiber, once the technology is better understood?
It doesn't really matter where we discuss this topic, in the end, since overlapping subjects usually get linked back and forth between the two forums (LMT and FCTF) in uncanny stereophonic fashion, borne of well-intentioned last-mile investors and enthusiasts alike, anyway.
======
Regards, Frank Coluccio |
