February 7, 1997

TOWARD A SUPERNET

by Andy Oram
American Reporter Correspondent

CAMBRIDGE, MASS.—Most Internet users take for granted the immense, complex telecommunications structure that underlies their Web surfing. But we can no longer afford blithe ignorance: The Federal Communications Commission (FCC) is considering telephone rate changes that could drastically raise the costs of using the Internet.

Although current hints suggest they will do nothing right now to affect Internet costs, the issue is clearly on the agenda and Americans should take an interest in it.

The well-publicized America Online bottlenecks are just a sideshow to this main-ring event. Whether a particular access provider (an Internet Service Provider, or ISP) has enough routers and lines to serve its customers is a trivial question, solved by buying more hardware.

The future of the Internet revolves around other issues: the lines used by individuals to call up their providers, the local and long-distance network used by Internet providers to connect with each other, and the whole technical structure of the voice-oriented phone network.

The FCC, industries that want to provide Internet access, and the public interest community (people like me) are all seeking answers.

The first indication that Internet usage may no longer be charged in the way it has been—at the price of a local call for many city dwellers—has been up to now was a rather idiosyncratic petition received by the FCC in March, 1995.

The petition was submitted by an obscure organization called America’s Carriers Telecommunication Association (ACTA), which represents smaller long-distance companies known as resellers, because they buy access and equipment from major carriers. The petition pointed out that the long-distance companies have to pay high charges to local phone companies in order to deliver calls across the local networks.

The ACTA petition complained that ISPs don’t have to pay such charges, even though, obviously, they deliver data around the country and the world. ISPs were explicitly exempted by the FCC in the early 1980s because they were considered “advanced services”; in short, the FCC said the Internet was a worthy experiment that should be given room to develop.

The details of the ACTA petition were quickly refuted, and the petition was turned down by the FCC. But phone companies were waking up to what computer people had said for years: digital networks would start merging with traditional media like the telephone and television. What bothered ACTA, in fact, was that some people were making free phone calls over the Internet using cheap, readily downloadable software such as IPhone, IRCphone, NetPhone and the like, rather than using the long-distance network directly.

The Internet averted crippling regulation that time, but the matter was not put to rest. In the fall of 1996 several local phone companies—notably Pacific Bell and Bell Atlantic—started complaining about households that dial up their Internet providers and stay on the telephone a lot longer than traditional voice calls. The companies’ investment in equipment—and thus their charge for local phone service—is based on the average number of calls placed and the length of those calls; the Bells claimed increased use for the Internet was throwing off their whole pricing system.

Their proposed solution was to impose metered rates (that is, paying per minute instead of per month) on ISPs. Presumably, the ISPs would pass these costs on to consumers and start to meter usage. In fact, the policy that got America Online in trouble—offer unlimited usage for $19.95 to compete with smaller providers—came about because customers chafed under the yoke of metering, which, if it became the norm, would kill the open spirit of exploration that gives the Internet its greatest appeal.

Many critics have picked apart phone company claims, alleging they fudge their statistics (generally, since no one can collect such statistics except the companies providing the service, it is a hard claim to support), found flaws in their reasoning, and cynically pointed out that the companies complaining about Internet providers were entering the same business themselves—could their claims be a way to saddle their competition with more costs? Overall, Internet use is still such a tiny part of telephone use that it is hard to imagine it placing a noticeable burden on the telephone network.

Since a change in phone charges would have to win FCC approval, the discussion has entered the public arena. The FCC held an all-day forum with company and public-interest representatives in January 1997 to explore all the options. While access charges for the Internet are still on the table for discussion, the FCC seems to be indicating that they won’t do anything to cause such an upheaval in service right now.

They may, however, make other changes in the complex rate structure to satisfy the politically powerful phone companies. Other companies—particularly those making computer hardware and software—counterbalance those demands and so far seem to be maintaining the status quo for Internet users.

Whatever the truth, there is no denying that in the future the Internet will grow to swallow more and more bandwidth. People who hope to get most Americans access to the Net—as demonstrated by the President’s State of the Union call to put all schools and hospitals online, and to build yet a second Internet—have to build a network that will support such “universal access.”

That new network has to be appropriate for data transfer. The phone companies say they would like to do that, but the cost may be as much as $50 billion, although new assistance from the military—such as the Army Corp of Engineers—could bring that price down. Such a network would use packet-switching (sharing lines among many users and sending their data in chopped-up portions, as the Internet itself does) instead of circuit-switching (the perennial telephone technology that dedicates a line to each two people trying to make a call).

The Supernet, if we may coin a phrase, would use much higher-speed optical fiber arranged in thick cables called C12 lines that would move in a few pulses of light what a 28.8 baud modem would take several hours to transfer. With that kind of muscle in the Internet backbone, Americans could face the computerized future with solid confidence.

As an alternative, the Internet could be carried over over cable TV, or best of all (in theory) over wireless networks that use cellular equipment, satellites, and broadcast towers. These alternatives still face technical problems, though.

So it is still impossible to predict with confidence what the Internet of the 21st Century will look like. But we had all better be ready for change. Many companies are busy building technologies that depend on keeping current regulatory policy.

For instance, there’s growing interest in “Virtual Private Networks,” which simulate the efficiency and security of a private corporate network of dedicated lines by routing traffic over public lines. A change in Internet charges could remove all justification for building these “poor man’s internets.”

Ultimately, the public will pay for the Internet of the 21st Century. And the public has a right to decide how. Will we pay through income taxes, everyday phone charges, user fees, content fees, or what? We can’t just say, “Leave it up to the market,” because the telephone and television networks are already enmeshed in a web of regulations. And those private interests may not want to fulfill the public interest in rapidly developing a powerful network that can support a hundred millions users who would be freed to work from their homes.

And if we do it wrong? New tariffs could leave some companies, such as the local phone monopolies, with unfair advantages. Rates could favor the easiest-served markets and exclude poor or rural users. Different companies may find it impossible to interconnect, so that users cannot talk to each other.

Worst of all, we might betray the promise of all the technology unleashed by the Internet, and the promised the future might not come at all.

Regulatory agencies must work hard to avoid all these problems and deliver the goods to a future America where labor is done at a keyboard rather than an assembly line.


Editor, O’Reilly Media
Author’s home page
Other articles in chronological order
Index to other articles