Sunday, August 11, 2013

Adapting Old-Style Phone Wires for Superfast Internet

Alcatel-Lucent has demonstrated fiber-like data-transfer speeds over telephone wiring—but will ISPs adopt it?

New technology can blast gigabit-per-second data speeds across age-old twisted-pair copper telephone cables—at least at distances from a telephone pole to a house, says Alcatel-Lucent.

In theory, such technology could be crucial to speeding up global Internet access. Of the 580 million broadband subscribers in the world, 55 percent have copper connections—though that figure is 33 percent in the United States, where most people get their broadband from the same coaxial cable that delivers their TV, according to Dell’oro, a telecommunications market research firm.

Fiber-optic cable, of course, beats coaxial cable, but installing it is costly. The new Alcatel-Lucent technology could bring fiber-like speeds from the fiber running along streets to each home—without requiring fiber installation along this final stretch.

“Fiber-to-the-home is obviously the most suitable technology, but you are installing new infrastructure, which requires a new installation in every home. If you want to cover an entire country, it will take 10 or 20 years to do that,” says Stefaan Vanhastel, director of marketing for Alcatel-Lucent’s fixed networks business.

The copper turbo-charging technology developed by Alcatel-Lucent uses an upcoming standard known as G.fast that got a preliminary step toward approval two weeks ago from the International Telecommunications Union (ITU). The technology involves using a much greater range of frequencies on copper cables. Because copper is very prone to interference, or “cross talk,” between nearby lines, this is continuously measured—and a noise-canceling signal is generated to counteract it.

G.fast could eventually be deployed on a telephone pole, a wall, or in a manhole. A shoebox-size device could contain the necessary electronics; a fiber terminus would be connected to a dozen or more existing copper lines leading to nearby houses. Alternatively, it could be installed in an apartment building basement, if a fiber terminus was already there, to jack up speeds for all the apartments via the telephone cables. The ITU is expected to approve the final standard in 2014, and products are expected to hit the market after 2015.

While the technology itself may be impressive, the major roadblock will be convincing carriers to spend anything to upgrade the networks, says Blair Levin, a former chief of staff at the U.S. Federal Communications Commission. Levin cowrote the National Broadband Plan and is now executive director of Gig.U, a consortium of universities working on installing fast networks.

Levin says there is little incentive for ISPs to provide a better service because they’re already so profitable. “Efforts to cheaply and massively upgrade copper always reminds me of Pope’s lines: Hope springs eternal/in the human breast,” Levin says. “But when major telcos deploy this and it works, I will be delighted to find a new poem.”

Some current systems use similar approaches to achieve 300 megabits per second in lab tests, and 40 to 60 megabits in the field, say Michael Timmers, who helped develop the technology at Alcatel-Lucent’s Bell Labs (see “Achieving Fiber-Optic Speeds over Copper Lines”). In field tests with Telecom Austria, the new technology hit 1.1 gigabits per second at a distance of 200 feet, and 500 megabits at 300 feet.

Other approaches that increase the speed at which data can be sent over copper include improving the efficiency of data encoding and frequency modulation (see “A Bandwidth Breakthrough”).

“It will be expensive to replace the installed base and, hence, quite worthwhile to squeeze all the performance that can be obtained,” says Muriel Medard, a professor at MIT’s Research Laboratory of Electronics. “There are several approaches, generally based on modulation—but also coding [of information] that will allow better use of the copper resources and delay the need for onerous upgrades.”

But wherever the improvements come from, they will still require the Internet service providers to care about improving those speeds. Levin says the factor most likely to alter market dynamics is competition from the likes of Google Fiber—Google’s effort to install cheap one-gigabit-per-second service in Kansas City (see “When Will the Rest of Us Get Google Fiber?”) and elsewhere.

In similar efforts, some municipalities are providing ultra-cheap, ultra-fast speeds; one such city is Chattanooga, Tennessee, whose local power utility used federal stimulus money to build a one-gigabit network (see “City with Superfast Internet Invites Innovators to Play”).

http://www.technologyreview.com/news/517381/adapting-old-style-phone-wires-for-superfast-internet/


No comments:

Post a Comment