Today's tech is a bit of a departure for us. It's not a review or a do-it-yourself project, but rather, a report on a DIY project that we hope will enlighten and inspire Street Tech readers. This is the first in a series of such reports on innovative, low-cost ways of using computer, Internet, and wireless technologies in ways that their creators might not have expected...in other words...consummate street tech. Starting with this piece, we're also offering links to a Street Tech Glossary for those readers who might like being brought up to speed on some of the geekspeak. The glossary will grow with each review and article. As always, let us know how we're doing, and if we're moving in the right direction, in our Shop Talk conferencing area.
- Gareth

For those seeking high-speed Net access, few demons loom as large as the keepers of the local loop. Charging upwards of a thousand bucks per year for a dedicated ISDN line (installation not included), the local telcos effectively apportion fat datapipes only to the rich and urban. Luck and your friendly cable company might bring cheaper high-speed access in the coming years, but you're still looking at regular charges that could top half a grand per year.

For those who lack the funds or proximity to afford such luxuries, Net access has remained a matter of a modem and a POTS line -- meager tools indeed for the dawning age of multimedia networking.

Enter Dave Hughes, a seventy-year old retired Army colonel who, along with a team of DIY techies, has set out to prove that a bit of ingenuity, some WWII-era technology, and a short stack of cash can be combined to give "the toughest link in rural Colorado" high-speed access to the Net.

"The public assumes that the only way to connect is through the phone companies," says Hughes. "To tell people that you can have high-speed, secure, recurrent-cost-free bandwidth wherever you live...well, it just causes a mental block in most cases. People don't believe it. And it's certainly in the interests of the phone companies that you don't know."

One of the Net's most recognized elder statesmen (he's been profiled in the Wall Street Journal, Wired, and countless other publications), Hughes became concerned with the issue of affordable access almost as soon as he arrived on the Net in the late '70s.

"I saw very early that there was going to be an information-rich/ information-poor split coming, not so much in the urban areas, but in the rural areas where the cost of communications is so high," he says.

Hughes soon became involved in a project aimed at offering networking to schoolteachersnd-forward technology, utilizing UUCP. "It was an economical way to give connectivity at the grass-roots level to rural America," he says.

But that was the '80s. By the early '90s, "that wasn't going to cut it for the Internet," Hughes says. "You couldn't do store-and-forward for the Internet, you had to be continually connected. So I started futzing around with what's called packet radio."

Spread: The Gospel

But the problem with packet radio was speed. "It could only be as wide as 9600 baud, because of the bandwidth of frequencies that were allocated." So Hughes began looking at a technology known as spread spectrum radio.

Spread spectrum radio is theoretically simple. Two radios communicate by hopping from frequency to frequency, transmitting and receiving packetized data. "Originally, it was created during World War II for security reasons," explains Hughes. "If you take the radio signal, and you have some way to stuff it through a whole bunch of different frequencies in sequence, then it's almost impossible to intercept or jam...The radios have IDs built into the chips, so you have to tell radio A that it's talking to radio B specifically. Then they're synchronized, and they have to exchange all sorts of signalling...Even if you have the same radios made by the same manufacturers on the same frequency range, you're not in the loop, because it's a dialog back and forth. So, without encryption, this is already secure communication."

But spread spectrum's security is not its only strength. "If you do it right, it doesn't interfere with other radios," Hughes explains. "And if you don't interfere with other radios, then you don't need a license. And if you don't need a license, then technically today, you could have a million spread spectrum radios in New York City, for instance, and they wouldn't interfere with one another."

What's more, spread spectrum radio has virtually unlimited bandwidth capability. Despite limitations on the frequencies and wattage available for public spread spectrum use, T-1-speed radio links are already possible, and researchers are aiming to create affordable radios that will offer Ethernet-level bandwidth.

What's the catch? Range. Because the FCC has limited unlicensed consumer spread spectrum radios to one watt of power, and because the frequencies available are fairly high in the spectrum, the spread spectrum radios available to the public have difficulty communicating through more than a few walls. Even with direct line-of-sight between antennae, the radios are limited to a maximum range of about thirty miles.

Nonetheless, Hughes believed that a bit of ingenuity could make up for the limitations imposed on the technology by the FCC. To prove it, he decided to build a high-speed radio link to the remote town of San Luis, Colorado.

One of the oldest non-native settlements in the American west, San Luis is home to a mostly Hispanic population and a K-12 school that serves fewer than 400 children. Until Hughes came along, the school's technological capabilities were limited to fifteen computers hooked to a 286, Token Ring-based Novell network in the business classroom. Net access required a long-distance call to the closest town, Alamosa, some thirty miles away. Even a single ISDN line connecting the school to the Net was prohibitively expensive for the small school.

Because it is situated in a valley and across a mountain range from the closest Net node, San Luis provided the perfect challenge for Hughes and his team of wireless evangelists. "If we could prove this technology could offer a high-speed link to San Luis, then we could prove the same technology would work virtually anywhere," says Hughes.

Armed with a modest grant from the National Science Foundation, Hughes and his team purchased four Freewave spread spectrum radios, five routers, four antennae, a spool of RF (radio frequency) cable and various tools. One radio was placed in the office of Amigo.net, an ISP in Alamosa, CO, and connected to the Net by way of three Debian Linux routers. A sixty-foot radio tower topped by a 3' yagi antenna was built on top of the Amigo Net office and connected to the radio. The antenna was pointed across the Rio Grande and between two hills, toward a Sheriff's Department radio tower some 30.2 miles away in San Luis. With only 10' of clearance over the hills, the margin of error for the one-watt transmission would be slim.

At the Sheriff's Department in San Luis, a second yagi antenna was installed atop the tower, and connected to the second radio in the shack below (a step which was hampered for two weeks earlier this year when no one could find the key to the shack). A 286 MS-DOS 640Kb router connected the second radio to a third, which transmitted its signal via an 8' omni antenna on the roof of the shack to a fourth radio in the communications room of the school in the valley below.

Finally, two weeks ago, the entire system was in place. "Appropriately enough, we ended the final tinkering after dark, in the radio shack with nothing but flashlights, in a chill wind, up on the bleak ridge, damned near getting stuck when we high centered a rock with 3 hours of driving still to do," Hughes says. The next day, for the first time, students in San Luis were able to access the Internet at speeds up to 115Kbps. And with the radio relay in place, the school will never have to pay the local loop one thin dime.

Bringing it home

The radios themselves cost a pretty penny, with the Freewaves retailing around $1,200 each. But considering that an ISDN line will cost in excess of a grand per year (and that's assuming your local telco offers ISDN in the first place), Hughes believes these radios represent a cost-effective solution to institutions and individuals alike.

"The Freewave radio is probably the most universally interesting and useful one for people, because it has a serial connection that'll simply plug into the back of your PC...They also require extremely low power; you're talking about 1 watt, and 500 milliamps. You can plug it into a cigarette lighter or a two foot by one foot solar panel. So you can use it as an endpoint, or as a stationary relay that doesn't have to be along the power grid."

Hughes is certainly not the first to notice the potential of spread spectrum. Increasingly, ISPs such as Seattle's Metrocom are scooping up radios and leasing them to customers. Indeed, Hughes says that even telcos are getting into the action. "Behind the scenes, U.S. West is buying Sci-Link radios, which are T-1 channelized (which is necessary for voice traffic), to escape the local loop themselves in rural areas," Hughes says. "They're spending $10,000 each for these radios that they turn around and sell to you as cellphone."

And if that doesn't convince you to throw down some cash today, Hughes points out that TAPR is now in the final stages of developing a 512Kbps radio "that'll cost under $500, and it'll be in the...no-license, general-use DIY spectrum."

At that point, the local loop's stranglehold can finally be unraveled by anyone with a credit card and a brain. "I maintain that the only real cost of the information age is the learning curve," says Hughes. And he'd love nothing better than to school the world in how to beat the Man.

"I'm not interested in creating a dependent relationship in the work I do. I've only succeeded when I reach a point that they don't need me anymore."

- Joe Nickell [11/13/97]