Hi Reb, thanks for the kind words. You ask,
" I ask because I can envision someone like Global Crossing setting up a station (much like a drilling rig) in international waters for transmission to avoid landing fees and such that a cable would incur."
I was going to say a few things concerning why not, but then thought better of it, and decided to think about this one for a moment. I can think of several show stoppers. While offshore drilling stations are loaded to the brim with satellite earth station and line of site microwave and other types of atmospheric radio gear, ordinary I-R line of sight (LOS) systems are not capable of anywhere near the capacity that GBLX would need to entertain at a cable landing. Perhaps these devices could serve some better purpose crossing mountain ranges (thinking out loud for a moment), or getting from one side of a canyon to the other (really thinking out loud now) in order to avoid trenching in some extremely hostile environments. But let's remember that these devices are designed for very short distances, and the higher the bit rate, the shorter the distances that are attainable.
If we assume that it is possible to do what you envision, consider that if a hosting shoreline nation was hostile enough to warrant exclusion, then getting the shot onto shore by I-R would be a Pyrrhic victory at best, because that nation would not permit them onto their PSTN, in any event.
But it's not entirely implausible that they would find some utility for this kind of arrangement, even if for only temporary measures.
The following two performance areas, however, would constitute the most formidable show stoppers: 1. Capacity Constraints; and, 2. Reliability and Availability
1. Capacity Constraints
Even the LU model which we discussed yesterday (the OpticAir device) can only carry up to 10Gb/s over four (4) fully loaded wavelengths over their IR system. Personally, I think that that is pushing it a bit, but we'll go with the press release for the moment.
The unit, when equipped with DWDMs at each end, processes four wavelengths, each at 2.5 Gb/s. That's the equivalent of what one wavelength can handle on the submarine and terrestrial sides of the I-R (which would be placed between the ocean and the land sections), and there could be up to 40 or more wavelengths in this current generation of DWDMs on both sides of the laser shot (i.e. onthe ocean and land sides), contained within each individual optical strand. Forty wavelengths, or lambdas, for the purposes of this discussion.
(The current systems may only be using 8, 16 or 32 lambdas per strand, but future systems will use beyond the 40 I mentioned, so let's stay with the number 40 for now.)
If we assign some nominal throughput capability to the fiber sections (submarine and land) of, say, 800 Gb/s (2 basic multiplexed strands * 40 lambdas/strand * 10 Gb/s/lambda), then this would equate to a total of one hundred and sixty (160) individual I-R transceivers, or 80 at sea and 80 on land.
This would be highly cost prohibitive, administratively and logistically untenable, and at the same time, a windfall for LU. -g-
2. Reliability and Availability
Fiber optic systems have been known to pass data unobstructed for days, weeks, and even months without ever incurring a single measurable bit error. I-R systems, on the other hand, are rated for their average weather days at 10^-10 or 11, under ideal conditions, during which times they perform better than outdoor copper, but worse than fiber. Those are on average weather days, assuming average weather is clement.
Given even a marginal degree of atmospheric anomaly, the error rate begins to increase proportionately. Minor rainfalls could cause real problems, while fog could take them out altogether. And storms could cause error performance to deteriorate to the extent that these systems are totally unusable for sustained periods of time. Given the nature of present day service level agreements (SLAs), these outages could result in large rebates, and in many cases contractual penalties, for missed minutes, hours, and days of use. It would only take one or two of these outages before GBLX would have a hard time drawing flies.
No, I don't think that this is the application they had in mind when signing up to be the first beta customer.
That one still has me scratching my head, scratch.. scratch.. , because even assuming that they wanted to use these units for customer handoffs in urban settings, the same corporate and carrier-level SLAs would apply in the Canyons of NY City as they would off the coast of Libya. Anyone have any suggestions as to why GBLX may be interested in such short-haul and highly unreliable (comparatively speaking) platforms?
Regards, Frank Coluccio |
