A few weeks ago I posted a series on the road to 100G, and one of the points I made was that there is an industry bias against bandwidth aggregation. Well, here is a case in point. The online magazine xChange reports that one ‘prominent industry analyst’ called it “Infinera’s fake 100G demo” because it was done via bandwidth aggregation.
In a way, I understand this sentiment, because for some fiber-poor networks 100G via bandwidth aggregation doesn’t solve one of the problems they need to solve, which is to fit more and more data onto the fiber they have. But for fiber-rich networks out there, the bigger issue is in how costs will scale going forward. It still costs more for a 40G pipe than it does for four 10G pipes. You can claim savings on the simplification of operations that justify it, but it doesn’t change the fact that the cost per bit is still of the same order of magnitude as bandwidth grows 50%+ per year.
Sooner or later the industry is going to have to face the mathematics. Solving the technical hurdles is not enough, putting 100G on a single wavelength is just not likely to be able to keep pace on cost for a long time. Fake or not, bandwidth aggregation is not going away.
If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!
Categories: Internet Backbones · Telecom Equipment
Join Our LinkedIn Group
I wonder if the “prominent industry analyst” has bothered to read the IEEE HSSG drafts and meeting minutes? All the IEEE interfaces for 100GE so far will be based on some form of b/w aggregation.
You could argue that “native” 100G wave implementations are all based on bandwidth aggregation. It’s just that they aggregate different phases, or polarization states, or both.
Sounds like this guys been to one too many vendor hospitality events.
hR.
“Sooner or later the industry is going to have to face the mathematics. ”
Yes, or so it would seem. In actuality, however, the industry has been steadily solving these same mathematics for over a hundred years. First with individual wireline circuits being modulated onto an open wire carrier system atop poles; then telegraph channels aggregated onto a single voice-grade channel on submarine channels; then analog carrier groups employing AM/FDM aggregated onto yet larger carrier groups and ultimately super- and jumbo- groups supported by coaxial “tubes”; and then came microwave and then fiber.
I’ve not yet seen anything in the optical realm that t doesn’t already have an analogue in past multiplexing schemes, save, perhaps, the impediments introduced by polarization mode dispersion, and soon the need to deal with issues surrounding the speed-of-light constraint in planars, and even fiber, itself.