By John Borland
Staff Writer, CNET News.com
February 4, 2003, 4:00 AM PT

Scrambling atop the roof of an Oakland, Calif., office building, John Furrier squinted like a surveyor from behind his company's wireless antenna, trying to figure out exactly where it pointed.

The Etherlinx co-founder was drawing a bead on a distant office that used his start-up's technology, receiving a fast Internet connection from a transmitter the size of a spiral notebook. As a pair of prospective clients awaited an ad hoc demo, he told them that they too could create their own broadband ISPs, competing against the phone companies with barely any start-up costs.

"Telephone companies and the other big companies are like a big rock in the river," Furrier said. "We could wait for it to move--or we can let the water flow around it."

A start-up with as yet more buzz and ambition than hard cash, Etherlinx is part of an expanding second generation of wireless companies betting on new and inexpensive technologies to help them avoid the mistakes of their predecessors. The most ambitious among them talk of undermining the business of telecommunications companies thousands of times their size.

They are part of a burgeoning wireless movement that bears remarkable similarities to the entrepreneurial garage culture that birthed Apple Computer and other Silicon Valley icons. Where those computing pioneers labored in the shadow of massive IBM, these wireless tinkerers are challenging the cable and phone companies that control the lines into most homes.

The key to their success, according to Furrier and others, is filling the gaps in vast networks constructed by large communications companies. At least today, their tools of choice are variations of the wireless technologies known as Wi-Fi or 802.11.

The cable and DSL companies control access to their networks, making it increasingly difficult or expensive for independent ISPs to offer broadband service. By contrast, wireless supporters say their technology is cheap, easy to set up, and operates in parts of the spectrum that are free to the public--unlike the walled-off portions of the airwaves that cellular phone corporations purchase for billions of dollars.

"The opportunities are pretty significant," said John Yunker, an analyst with Pyramid Research and editor of a wireless newsletter. "Wi-Fi really marks the beginning of the end of the wall between fixed network operators and mobile operators. All the lines are blurring."

It's not the first time wireless technologies have inspired revolutionary ambitions. At the height of the Internet bubble, wireless was routinely cited as the "third pipe" to the home. AT&T, Sprint and WorldCom, along with a myriad of start-ups, collectively invested billions in fixed wireless technology--as opposed to the mobile wireless technology used for cellular telephones--designed to bypass the lock held by cable and local phone companies on the wires into homes.

For the most part, this first generation fell flat. All three long-distance companies have decided to halt or stop expanding their wireless projects, saying the technology is not yet ready. Start-ups offering fixed wireless service have found it difficult to stay afloat in the telecommunications market's meltdown.

Not ready for prime time
Nevertheless, for all its enthusiasm and momentum, the grassroots Wi-Fi movement is hardly an automatic solution for providing basic home Internet access. Although it comes in a half-dozen varieties, the most popular Wi-Fi technology creates a "hot spot" of network connectivity only about 600 feet wide. Without technology that relays or amplifies the signal, service areas are relatively small.

Moreover, Wi-Fi technology itself isn't designed to accommodate multiple people logging on at once. Wireless experts say this can lead to a situation where requests trip over each other, interfering with transmissions.

"It's feasible, but there are a lot of limitations," said Roger Marks, chairman of the standards-setting body IEEE's (Institute of Electrical and Electronics Engineers) working group on 802.16, another set of wireless technologies. "As you get more users, (Wi-Fi) is going to bog down, and not just linearly. If you get twice as many users, throughput goes down more than twice."

A handful of commercial 802.11-based ISPs have sprung up, but most offer Internet access in small areas such as cafes or airports for people using laptop computers. Local wireless groups sprinkled across the country maintain lists of areas where free access points are available even for home connections, but surfers have to be within hot spots to get working connections.

A coalition of tech giants including IBM, AT&T and Intel formed a new company called Cometa in December, with the idea of creating a nationwide network of hot spots. That company will wholesale its service to hotels, cafes, ISPs and other companies.

A handful of companies are toying with "mesh networks," in which multiple Wi-Fi or other wireless access points in a grid could talk to each other and create interconnections that stretch far beyond a single receiver's range. The idea is a little like the peer-to-peer networks that have formed online for swapping music, movies and other files.

In a mesh network, each computer or access point could serve as a relay station, providing connections to another set of computers and so on. Eventually, the entire chain of relay points would have to be linked to a wellspring of Net connections such as a DSL or T-1 line, but the concept theoretically expands Wi-Fi's range substantially.

Florida-based Mesh Networks is one such company, with technology licensed from the U.S. military. Developers there modeled their technology after the Internet itself, with decentralized pieces of the network routing traffic in whatever way was most efficient at any given moment.

Furrier's company is taking a different tack. Etherlinx uses off-the-shelf 802.11 hardware and modifies it to work in an entirely different manner. The result is a wireless service that can stretch 20 miles instead of a few hundred feet and can be relayed from station to station in ways similar to a mesh network. One set of the antenna and radio hardware can support 94 subscribers, each with DSL-like speeds of 256kbps.

Anyone can be an ISP
The equipment is inexpensive to purchase and set up, making it possible for just about anyone to become an ISP as long as they have access to their own DSL or other broadband connection, Furrier says.

Etherlinx gained international attention after the start-up's work was cited in a New York Times article last year. The publicity didn't immediately translate to investment, and the company ultimately launched a search for venture capital to finance production and new features, however.

Grilled on the phone by a potential client interested in new features for the hardware his company is producing in small quantities, Furrier replied a little peevishly: "Anything's possible with money. We don't have a lot of money."

Without substantial resources, wireless start-ups are still far from making a dent in the broadband market dominated by the likes of Comcast and Verizon Communications.

Major companies are certainly interested in Wi-Fi, as Cometa's creation shows. But making it a viable substitute for DSL or cable is another matter. Verizon, which advocates use of the technology for home networking, says there are simply too many technical problems. Because the service runs in unlicensed spectrum--meaning anybody can start services that interfere with each other--it would be impossible to guarantee reliability at this stage.

"In that respect, too much success is likely to breed failure," said Brian Whitton, Verizon's executive director of network evolution. "It's not conducive for a network-based offering."

The IEEE's Marks said the next generation of the technologies already tried by Sprint and other fixed-wireless companies will be better suited for home broadband connections. These technologies will be included in the standard created by his 802.16 group.

A few industry analysts, including Pyramid's Yunker, are telling the large phone companies that they need to understand that the small companies and still-maturing technology can be a threat, however. The energy driving Wi-Fi and related technologies is building by the month, they say, and the large companies could wake up one day and find that real grassroots competition has emerged seemingly from nowhere.

"For them to fight it, they need to embrace it," Yunker said.