Enjoyed reading your description of the backbone cable assembly. Time is flying to be sure, and yesterday's awesome leaps in bandwidth capacity are becoming as tantalizing as having dinner at a greasy spoon.
>>The backbone is a half inch multiple stranded fiber optic cable(s) buried in protective pipes underground. <<
I keep going back to my childhood recalling the Million Dollar Movie on Channel 9 in NY City, a TV show during the fifties that presented the same movie three times every day, for a week. There were no infomercials then, and tummy stretchers weren't invented yet, so advertising was still a novelty dominated by the top brands of the day.
One of the movies I recall most vividly was the Hunchback of Notre Dame. Must have seen it twenty times in one week alone (they replayed it the same way several months later, and I repeated the novena then, too). And the part that I specifically recall most was a one-liner stating that "the pen is mightier than the sword."
Of course, I've come to hear this many times since, but each time I hear it I still revert back forty or more years to that scene that talked about the revolution.
So, where am I going with this? The number of strands in the fiber cable keeps going up. Perelli recently announced a cable that can fit 6 gross of strands, that's 864 single modes into a standard form factor of less than an inch outside diameter.
So? Well, and don't quote me on this, but I seem to recall that they then came out with one that does 1,152 strands in the same form factor. I/e., eight gross. And if I am mistaken about this, we can at least use it for discussion purposes.
Note: I actually tried to go back and verify this, but their web site is more attuned to sexy calendars than providing spec sheets on their products. Go to:
pirelli.com
... to see what I mean!
Three of these 1152 strand cables can typically fit into a standard 4" conduit. That equates to 3,456 strands per outer tube.
At the soon-to-be DWDM rate of 128 lambdae, that's on the order of 442,368 wavelengths per tube...
...and at 10 Gb/s nominal per wavelength, that equates to ~4,424 Tb/s in a single tube.
4,424 Terabits Per Second in a cross-sectioinally under-utilized, but compliant, four inch pipe, using three inner-ducts.
Then there are at least three or four pipes laid at one time upon trenching, if not a dozen or more. But we'll only have to use a single conduit here to make a point.
4,424 Terabits Per Second through the use of multiple *thousands* of DWDMs is what I would call brute force. Not elegant, not entirely intelligent, but the power of brute force, nonetheless.
Now, to do something with this to demonstrate the benefits to be derived from using an SR or equivalent approach, we would have to figure out what the physical implementation and management of these wavelengths would entail, how much it would take to administer, and then extrapolate from there what the actual costs are.
Anyone care to take a stab at this? Didn't think so. But I've done some heuristic analysis of this, and thrown more than a few napkins into the circular file, and take my word for it, it's a bear. And yet, this capacity is not overly far fetched, when you think of the possibilities that exist, and the eventual need for photons to flow as freely as electrons do today in our electric power utilities.
And let's not forget, that these are indeed backbone situations, still, so there is still only a token need for "distribution" at mid points at this time, the way these routes are currently being set up by the long haul carriers.
How many DWDM devices would it take to manage in excess of 442,368 wavelengths? Try it takes 6,912 DWDM muxes to handle this single route. 3,456 boxes in city A, and 3,456 boxes in City B. And if these cities are situated three thousand mile apart, how many fiber amplifier pumps are required? I rest my case for the moment.
But don't kid yourself.. it is still I who I am trying to convince here. Not you, necessarily. Thanks for listening...
|
