Gazettabyte has an interesting piece about where the next boost in bandwidth is going to come from now that 100G is making its way into production networks. They detail the efforts of a European research group, called MODE-GAP, looking for ways to multiply the total capacity of an optical fiber by another 100. If we can put 10Tbps on a single fiber now and perhaps 20-25Tbps in a few years, then what they're talking about is levels of 1 Petabit per second.
They're looking at getting there via combination of spatial modulation and a new beast (to me) called photonic bandgap fiber. New fiber eh? That will certainly shake things up at some point, given that so many networks are still getting by on stuff that's been in the ground for decades. (Yes, even Level 3's buildout was more than a decade ago now!)
The past decade or two of innovation have allowed us to put an amazing amount of bandwidth on a single fiber, so much so that despite the continuing exponential traffic growth most of the longhaul fiber that went in the ground in the late 90s is still dark. Having the right size pipes to build the internet with is important of course, but it seems to me that the coming 'traffic crunch' isn't really that threatening.
There's a fundamental design difference between fiber and, say, wireless spectrum. The spectrum we have now is all that we will ever have, but we can always add more fiber to a route if the price is right and the demand is there. How hard does it make sense to push the technological limits of bandwidth per fiber when we can just add more strands? Isn't the most important number not the number of bits per second, but rather the cost per bit per second?
For one thing, raw backbone traffic isn't really straining the limits of technology right now. It's not the bottleneck, nor has it been in a long time - but that's where a 1Pbps technology would be aimed at. The last mile and the aggregation layer are the cost centers that are giving service providers the willies when designing their networks. That's where the money is to be made and hence where the research money is best spent. You don't need the very largest pipes there, just cost effective medium sized ones.
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 Traffic · Telecom Equipment