TCS Daily


By Duane D. Freese - February 8, 2002 12:00 AM

Forget about broadband. Think gigaband. Or rather GIGABAND.

That, at any rate, is what Douglas Van Houweling hopes policymakers will start doing.

At a New America Foundation forum late last month, the CEO and president of Internet2, a consortium of 190 universities and some 70 business partners, including AT&T, 3Com, Qwest and SBC Research, outlined the problems and the potential of superfast connections of a new Internet.

What is GIGABAND? Technically, it's a delivery speed - a billion bits per second. That's 30,000 times faster than the 33 kilobits per second rate that the Internet delivered to residences in 1995. It's 1,000 times faster the 1 Megabit (1 million bits) per second rate that broadband available at universities back then delivered.

Considering the crawl to old, slow broadband - only 10 percent of homes have it now - is Van Houweling's vision a pipe dream? Not totally. In fact, Internet2, the development of which began in earnest in 1999, operates at the warp rate of 2.5 billion bits per second, 2.5 Gigabits, and is in the midst of an upgrade using new optical equipment that will carry it to 10 Gigabits per second in the next couple years.

But what would GIGABAND accomplish, really? Well, if made universally available, it would achieve the goal outlined by the by the high tech executives of TechNet and the Computer Systems Policy Project to the White House and Congress in recent weeks - 100 Mbps of download and upload speed, or approximately 200 times the current rate most residential computer customers get with their dial up service and 10 to 40 times people with current broadband.

Such speeds provide real time delivery of visual information without the herky-jerky motion of today's streaming video or the breakup that occurs on most television screens getting digital pictures. Indeed, at its most advanced, GIGABAND would make virtual reality - three-dimensional delivery of visual information, or tele-immersion - a reality. High-definition television plus, plus, plus.

With GIGABAND, an astronomer in Massachusetts can drink in the view from the Mauna Kea Observatory without having to go to Hawaii, spend time acclimating to the mile-plus high heights, and then have the trip possibly wasted because of cloud cover. A surgeon in Alaska can collaborate with a specialist in Florida advising every move of a complicated heart or other procedure. Rather than sharing specs by FedEx, engineers can examine machines together at the same instant even if they are thousand of miles from each other and the machines.

The possible uses of such technology to improve lives and even save them, and increase the spread of knowledge, wealth and health are incalculable. No single killer application for use but possibly dozens and hundreds and even thousands of uses that today we have no idea about, according to Van Houweling.

What stands in the way of GIGABAND's and Internet2's diffusion to small businesses and homes? Two things: Physics and politics.

Both problems have to do with reciprocal delivery and receipt of information over the last mile of pipe into the home.

Van Houweling noted that twisted copper pair phone lines that are being used to provide digital subscriber line (DSL) service at up to about 1.5 Mbps require an overlay of fiber optic lines to the curbside of homes.

Cable TV lines are better, he said, and many cable companies have upgraded their systems with fiber optic lines to the home. But the current sharing of bandwidth between users and services won't allow Gigabit delivery to a home. Restructuring would allow homes to dedicate 15 to 20 channels of cable - at 6 Mhz each - to provide real time service.

As for wireless, it would require a huge reallocation of broadcast spectrum - again about 20 TV channels worth - to allow a home to run at GIGABAND, Van Houweling said.

To allow for this means changing the conditions under which a lot of existing companies operate. That means overcoming their resistance to change.

New American Foundation Markle Fellow James Snider provided his own Schumpeterian vision for accomplishing that. He advocates the creative destruction of the old copper wire phone to make way for faster fiber lines and the old line broadcasting networks to make way for new wireless technologies.

Unfortunately, both the old phone and the old broadcasting networks, he noted, are too politically well protected.

The current digital television standard is already obsolete, and no amount of tinkering with it by Congress will make it better, he said. But Congress and local governments are wedded to localism of spectrum making reallocation of spectrum a slow process, he said.

On the wired front, he said that the outmoded nature of the current copper network is a potent argument for structural separation of each of the local Bell telephone monopolies into a wholesale arm running the old copper system and a new retail operation competing in new high-speed communication services. That would spur the spread of advanced broadband technology.

But Congress, he said, will protect the phone companies so they can continue to depreciate their copper plant. That's what makes many in Washington consider structural separation a nonstarter, despite the potent economic arguments for it from both conservative-libertarians such as Tech Central Station's James Glassman and liberals such as Stanford Law School Professor Lawrence Lessig.

Another NAF Fellow, Karen Kornbluh said that she, like Snider, was divided between the idealism of the possibilities of GIGABAND technology and pessimism about the politics of its rollout.

She noted that it was the networking of computers over the Internet that boosted productivity from 1.5 percent a year to above 2.5 percent in the late 1990s. She noted that the phone monopolies - when Ma Bell controlled the whole national network, nor under the Baby Bell's regional regimes created with AT&T's breakup 18 years ago - have done little to improve quality of service.

Fiber optic technology has been around for 30 years, but it took Sprint's pin drop ads and long distance competition to get fiber optic spread around the country, she noted. And it took spending of $90 billion to the phone system's backbone by competitors to the local Bells to get DSL broadband going to the homes, she said.

Kornbluh argued there are three approaches to promote rapid deployment of true high speed networks: The old Ma Bell route, with a promised return on investment for regulated monopolies; the highway model, with the federal government making the investment; and the structural separation route, which would allow competition to build out new networks. A big issue Congress must deal with is its continued demand for universal service that has cross subsidies to provide rural telephone service at below cost. Subsidies should be need based, she said.

Political intransigence and its support for existing providers will slow down the roll out of GIGABAND, but it likely won't stop it. Van Houweling believes people will demand those services. As university and even high school and even junior high school students who have access to GIGABAND find they can't get the same service at home and work, they will demand it, putting pressure on lawmakers to change, he said.

But how long will that take? His estimate: 8 to 10 years. Where's GIGABAND when you need it?


TCS Daily Archives