BROADBAND COMMUNITIES is the leading source of information on digital and broadband technologies for buildings and communities. Our editorial aims to accelerate the deployment of Fiber-To-The-Home and Fiber-To-The-Premises.
Issue link: http://bbcmag.epubxp.com/i/374665
6 | BROADBAND COMMUNITIES | www.broadbandcommunities.com | AUGUST/SEPTEMBER 2014 BANDWIDTH HAWK E ight years ago, fber-to-the-home pioneer John George of OFS wrote a long article for this magazine. In it, he predicted that homes would need at least 3 Gbps of bandwidth by 2030 and maybe as much as 30 Gbps. His projection for 2020 was 558 Mbps to 2.2 Gbps. At the time, the Federal Communications Commission defned broadband as 200 kilobits per second. A Gbps is a 5,000-fold increase. George got only one thing wrong: We're hitting the midpoint of his 2020 projection fve years ahead of schedule. Tat's despite the fact that he underestimated how good video compression could become. (Read the article in the magazine's archives at www.broadbandproperties. com/2006issues/sep06issues/george_sep.pdf.) Te wild card, of course, was the iPhone, introduced just after the article went to press. It went on sale in mid-2007, only through AT&T; within 18 months, AT&T's mobile data trafc had increased 50-fold. Historically, the volume of Internet trafc worldwide has been increasing at a long- term, remarkably steady 42 percent a year, roughly doubling every two years, thanks to compounding. By the end of 2013, however, trafc was already at the level that relentless doubling and redoubling had predicted for the end of 2020. Tat's seven years ahead of schedule, despite video compression! Only the fact that mobility spreads out peak trafc has kept the nation's networks from beginning to congeal. Two other developments are already beginning to increase trafc even more rapidly – the growth in uploading facilitated by local and regional data centers and, in the United States, the impact of a remarkably successful economic stimulus program. Although only $7 billion, or 1 percent of all stimulus funds, was spent on broadband and the spending came later than hoped for, it was for the most part brilliantly distributed. Te smart grid, FirstNET and the Internet of Tings still loom, and 50 billion machines are expected to be connected to the Internet by 2020. Middle-mile fber laid with the help of stimulus funds enabled gigabit builds in dozens of cities and towns nationwide. ADTR AN alone expects to provide equipment for 50 gigabit builds being announced by the end of this year and another 150 by the end of 2015. You don't have to partner with Google to get a gig. But is it hype? Even the technical community points out that the only app needing a gigabit is the app that confrms you have it. Tere's an answer for that: Every home is connecting dozens of devices, including multiple devices that multiple family members use at the same time. Bandwidth demand adds up fast. Te gigabit race refects ongoing competition between telcos and cable companies. Cable companies adopted DOCSIS 3.0 to deliver 25 to 30 Mbps (and in some cases much higher) and blew the telcos that were providing DSL at 5 Mbps out of the water. Telcos used FTTH and advanced versions of DSL to get back into the game. Cable Labs countered with DOCSIS 3.1, which has not yet been implemented. But telcos quickly pushed FTTH service to a gig without much extra expense (all the new middle-mile fber helped). Although DOCSIS 3.1 has a theoretical capacity of 10 Gbps downstream and 1 Gbps upstream, actually delivering these speeds using DOCSIS will require a great deal of capital investment and network redesign. Faced with the cost of splitting and resplitting DOCSIS nodes, every large cable company is at least considering conversion to passive optical networks or merger with operators who can aford to deploy PON. PON vendors all report quiet PON tests and limited adoptions by cable operators. Te biggest telco holdout, CenturyLink, has publicly committed to FTTH in larger cities in its sparsely populated footprint and to bringing as much bandwidth as possible, as fast as possible, to businesses. Fiber proftably run to cell sites has helped make the business case better in such areas. So the future is clear – telcos and cablecos eventually use about the same technology and ofer about the same services at a gigabit per second or faster. In places where they can't make a business case, local governments might – if citizens repeal the laws in 19 states that limit them from doing so. Tere is a bigger payof: economic development. Many businesses, even small ones, can beneft from that gig. Fortunately, fber is one of the cheapest, easiest-to-build categories of major infrastructure, and building for a gig adds little to the cost. Last year, a gig might have been hype. Tis year, it is the norm. v Contact the Bandwidth Hawk at email@example.com. Gigabit: Reality, Not Hype Last year, major carriers called gigabit services Google-inspired hype. Now, they're signing on. The impact may be greatest in rural areas. By Steven S. Ross / Broadband Communities