Spectrum for All

by Andy Oram
March 2, 2001

Wireless Internet is still in its infancy, and already we have twins. Two models are taking shape for getting the spectrum to support Internet access, and both exhibit strange social and business contradictions.

One model is being pushed by cellular phone companies: the well-known “3G” third generation or an enhanced 2G of mobile wireless, along with WAP and related protocols. This is the model that’s supposed to turn into a big pot of gold, and the Federal Communications Commission is rolling out gold-plated spectrum in big swaths to make the promise come true. But ironically, its business model is seriously flawed.

The other model consists of scrappy experimenters who seem to have no chance in the world of succeeding. They’re mostly using equipment that conforms to one of the IEEE 802.11 standards for transmitting data over short distances—a standard really meant just to be used within a small building or close-knit campus of buildings. And some of them are using fixed wireless as the basis for stable (although not thriving) ISP businesses: wireless ISPs or “WISPs.”

The pose taken by these small operators and community activists is one of ruthless independence and self-reliance. Describing the mutual benefit society he set up in Laramie, Wyoming to provide Internet access, Brett Glass describes the basic sociology as follows: “While people in Wyoming are very suspicious of government, they like working with people they know.”

Urban projects sound even more uncompromising. When I mentioned funding to Ken Caruso, who helps to organize Seattlewireless, he answered, “Seattlewireless is a loosely knit group of networking enthusiasts that does not have the facility to accept money.” To underline that this is a matter of principle rather than necessity, he added, “We want this network to be owned by the people who create it, and not have to report or be responsible to entities that would like to fund us.” They’re happy to let benefactors buy gear for individuals joining the network, though.

Caruso is the norm, not the exception. James Stevens of London’s consume.net says, “Funding nearly always sucks!…Working in collaboration without seeking financial gain has the capacity to add back some value into existence and undermine the motive to profit. It is also fun!”

The urban community networkers remind me of the early free software developers: people who were considered impossibly impractical and anarchistic (“we are not a formal organization by any stretch of the imagination,” says Stevens) but who yet carried out an apparently irreversible historical upheaval in the fields of software, information, and the Internet. The 802.11 networkers face many more barriers, however, than the free software movement.

For all their intransigence and refusal to deal with government, wireless Internet services are in desperate need of government intervention. The 802.11 standards use the areas of spectrum set aside in the United States for industrial, scientific, and medical (ISM) use, where interference can be injected by everything from microwave ovens and cordless phones to ham radio enthusiasts. “Trash spectrum,” says Allen Marsalis dismissively—but he bases a successful business on it.

Doug Moreen, a networking expert in Montana, says that 802.11 could work in cities, but that the equipment plus rent could come to $5,000 or even $10,000 for installation at each access point, plus $300 to $500 per customer. Cheaper equipment risks being overwhelmed by interference. Licensed devices are more powerful than the unlicensed 802.11 devices, and always take legal priority. The FCC has shown that it takes firm action against any ISP (an unlicensed operator by definition) that interferes with a licensed user such as an amateur radio operator.

So WISPs need new spectrum. They also need standards to make smart radios communicate in bursts without interfering with each other. The spectrum is a textbook case of the tragedy of the commons, where someone who chooses to blast away can jam everybody else. Finally, WISPs could use help from the universal service fund, which pays large telephone companies to wire schools but is effectively off-limits to wireless providers.

In the U.S., the FCC is the organization to solve some of these problems—and it was slowly making headway under Chairman Kennard. It will take a lot of pressure to bring the process to fruition under Chairman Powell.

The wireless pioneers

Allen Marsalis makes money by offering wireless Internet access in Louisiana. He uses several directional antennae in each location to transmit from one to three megabits per second over a distance of up to 10 miles, given a clear line of sight. He says cable and DSL don’t reach most of his region. “Many customers live five miles or more from the central office. Except for satellites, fixed wireless is the only hope.” A new report from the federal General Accounting Office finds that the percentage of Internet users in rural areas now matches that of urban areas—but that rural residents are more likely to be using slow dial-in lines and still lag in access to broadband.

Marsalis uses good engineering to make wireless work in a small urban area. But he knows his business model for broadband access is fragile. “We have wireless, so we can beat the Bells for the next 12 months,” he says. He would like to see the government provide better spectrum, but he possesses a thoroughgoing cynicism about its prospects. “The FCC and FTC don’t really care much for small local ISPs.” And while he’s in the same state (not the same district) as powerful representative Billy Tauzin, chair of the House subcommittee on Telecommunications, he can’t get Tauzin’s office to return his phone calls. He doesn’t get action from the state’s public utilities commission either.

Brett Glass created a completely community-run users group, LARIAT, in his Wyoming town to provide Internet service and related training. The group started using wireless because US West was overcharging for the landline services available at the time (DDS, dedicated T1, and Frame Relay). Since 1995, businesses could share their Internet connection through wireless, and with a new antenna they’re about to put up (renting the use of an old existing tower) they plan to offer wireless to individuals too.

“LARIAT is not idealistic,” says Glass, “it’s entirely pragmatic. People who are in there are there for their own benefit, doing things they couldn’t do on their own.” Legally, they’re organized as a non-profit mutual benefit society under Section 501(c)(12) of the Internal Revenue Code. “We don’t require altruism, just enlightened self-interest.”

One of the trickiest parts of offering wireless service is connecting to the Internet so that people can reach the outside world. Glass solved this through the same philosophy of mutual aid: he found an ISP that wanted a POP in Laramie. LARIAT provided free space for the ISP’s router and a place for it to terminate lines, and in return pays a reduced rate for sharing their T1 line to an Internet backbone.

Wireless can often work where customers are in tall buildings or near flat, open areas (including parking lots). Therefore, they are used by some intrepid communities to provide Internet access to schools. But FCC regulations make that harder too.

One of the most famous proponents of wireless internet is Dave Hughes of Old Colorado City Communications. For years he has pointed out that the definition of the government’s universal service fund excludes wireless, because it will pay only for “service offerings” and not equipment purchases. Because there are very few wireless services to choose from, and because the schools are blocked by the terms of the fund from buying the equipment to set up their own, wireless is not an option.

Moreen points out that it’s also hard for small providers to take advantage of government start-up grants, such as those offered by the Department of the Interior or Department of Commerce. The rules require extensive plans and evidence of community support, which is reasonable enough to ensure funds will be productively employed, but few small operators have the resources to meet the threshold for funding. Others have pointed out that, for the same reason, most projects getting reimbursed by the universal service fund are carried out by large, incumbent telephone companies.

Billions for private networks

Meanwhile, the FCC is disposing of unused spectrum by auctioning it to large telephone companies that might not even be able to use it. Michael Powell, the new chair, believes 3G is the route to the wireless Internet.

While the business commentators and phone companies are trying to excite us about the prospects of rolling into a new city and beaming ourselves right to a four-star Thai restaurant, European experience shows that Web surfers react tepidly to the marvels of a two-inch screen and telephone keypad. (Hint: having difficulty entering a URL? It may be easier to use an IP address instead of a domain name. Or pay for VeriSign’s new WebNum service.)

More and more analysts, therefore, hedge their bets regarding mobile wireless Internet—at least in the new 3G form—and phone companies are wondering when they’ll ever make a profit on the spectrum for which they just paid $17 billion.

In any case, the main role of the government is not to deal out scarce resources to private companies. A more appropriate role is to create benefits for the entire population in areas where private companies don’t know how to make a profit. The original Internet, which no private company would ever conceive of, is one such benefit. Use of the spectrum for Internet access is a natural follow-on.

What ever happened to the ideal of an Internet that was open to all? It doesn’t lie in cellular network gateways. “Privately owned spectrum for Internet access” is an oxymoron, like “local telephone competition” or “deregulatory ruling.” (But that’s what you can expect from people who throw around the term “wireless cable.”)

For the record, I have some faith that mobile wireless Internet access will eventually take off. But I think two pieces of technology will have to be in place before it’s widely popular:

Furthermore, the applications and experience of mobile wireless will remain different from fixed wireless; neither can be a substitute for the other.

A lagging initiative at the FCC

William Kennard, the recently departed chair of the FCC, took slow but deliberate steps in favor of a public wireless Internet. Thus says Dewayne Hendricks, CEO of Dandin Group (a wireless technology. initiatives, including the use of new technologies such as software-defined radios and ultrawide band. However, Hendricks cautions, “The FCC is under pressure to make progress, but nothing’s going to happen. The incumbents have a long history with the FCC and don’t want either competition or change in the more fundamental sense—innovation.”

If Powell really believes in deregulation (as contrasted with rhetoric that covers up price increases by monopoly providers) he should push as fast as possible to open up new unlicensed bands of spectrum. But he can’t just walk away from it once it’s allocated. Some regulation is needed to promote the development and use of smart devices that allow sharing by numerous competing companies. Glass says that one sharing system was tried by the FCC but that it proved inadequate. Further research is needed.

According to Hendricks, “There should be no new spectrum for unlicensed use until there are new rules.” He says that too many ISPS are already trying to share the spectrum in some places, and that the ISM band is so fragile that a single licensed device outside a Starbucks could drive everybody inside off the net.

Once the spectrum and the sharing algorithms are worked out, equipment is sure to follow. And then we will have a true third generation: a new generation, coming after the early Internet hackers and the dial-in Web surfers of the present day, of people and businesses who can use broadband Internet.


Andy Oram is an editor at O’Reilly Media and a member of Computer Professionals for Social Responsibility. This article represents his views only. It was originally published in the online magazine Web Review.