By Ron Hendrickson
All the signs are pointing to a coming boom in Internet traffic, so it would be wise to get ready for it. Is it time to rethink how much spectrum cable operators allocate to data?
The idea isn't new. European cable ops go up to 65 MHz for data, as opposed to only 42 MHz in the States, and there's been talk of going up to 85 MHz (as written into DOCSIS 3.0) both abroad and here. However, the idea is taking on new urgency.
Just yesterday, IDC issued a prediction that worldwide wireline and mobile broadband traffic will increase from 9,665 petabytes per month in 2010 to 116,539 petabytes per month in 2015. That's more than a full order of magnitude in just five years. Cisco's predictions track about the same.
One petabyte, by the way, is a million gigabytes (GB). In comparison, my laptop will hold 4 GB, so it would take 250,000 such laptops to hold just 1 petabyte. Now multiply that by 100,000 ... each month. Holy monkeys.
Much of the expected increase, both mobile and fixed, stems from increases in online video and audio streaming, peer-to-peer file-sharing, general web browsing, and the fact that pretty much anything that uses electricity is now connected to the Internet.
Short term, these are opportunities. Mobile backhaul is a huge vertical for cable ops, with a recent report by iGR predicting an 85% compound annual growth rate for fiber-based U.S. mobile backhaul between 2011 and 2016. And sales of CMTSs and DOCSIS 3.0 gear generally continue to be brisk. A couple weeks ago, Infonetics Research reported that North American CMTS ports grew 94% in 2011, and global CMTS port shipments hit record levels in 4Q11 and were up 48% for the full year.
Longer term, the question is whether just adding more CMTSs will be enough to keep up with demand. At the predicted growth rates, it probably won't. The IEEE is working on an EPON over coax (EPoC) standard that promises data rates up to 10 Gbps, but it's not expected to be finished until 2015 or so, with actual products to follow still later. So, what's to be done in the meantime?
One obvious answer is usage-based billing, which generally hasn't gone over well with consumers, but may become a necessary evil. Polish up your PR and complaint departments.
Another option is to rob bandwidth from the video space and repurpose it for data. Frankly, the vast majority of video programming on cable is crap anyway, and easily three-quarters of it could be dumped without losing anything of consequence. ("Hillbilly Hand-Fishing"? "Real Housewives of Wherever"? Please.)
It shouldn't be hard to clear out enough junk programming to at least double -- if not triple or quadruple -- the spectrum allocated to data. Doing so would definitely help cable ops to keep up with the continuing data surge, and perhaps also save some of the money currently spent on programming. Who knows, it may even improve programming across the board, as cable ops vote with their wallets.
What do you think? Should cable ops cut some deadwood out of the video spectrum so that data can grow? Tell us what you think in the comments section below, or feel free to e-mail me directly.
Ron Hendrickson is BTR's managing editor. Reach him at email@example.com.
Time to Rethink Data Spectrum? What Say Ye?
By Ron Hendrickson