What's Next

Disruptive technologies are those that change not just business but society. Personal computers threw the mainframe-computer business for a loop. Airlines replaced passenger trains. HMOs became an alternative to the family doctor. In all these cases, the computer, transportation, and health-care moguls of the time didn't view the new entrants as real competition, but rather as gimmicks, flukes, toys. And the new entrants were, in their early incarnation, inferior to the technology they replaced. But they appealed to some groups of consumers enough so that they came to compete with, and eventually surpass, the previous technologies. And this is what is happening right now with wireless networking, which promises to change forever the computing and communication industries, creating a lot of business opportunity as it does so.

Wireless networking in this context means strictly unlicensed digital communication using radio signals. That eliminates cellular and other mobile-data networks, which are typically owned by phone companies. What it includes is an unlicensed technology generally known as "WiFi," because people don't want to say "IEEE 802.11a, b, and g."

WiFi exists primarily thanks to Apple Computer, which years ago proposed to the Federal Communications Commission that certain radio frequencies in the 2.4 to 5.8 gigahertz range be allocated for unlicensed data communication. Today those frequencies are available for people to do pretty much anything as long as they stay below a certain power output and make nice-nice with the neighbors. Apple intended to use the frequencies for wireless local area networks and introduced its AirPort product line several years ago to do just that. AirPort allowed computers to link with one another within a radius of 50 meters, sharing data at speeds up to 11 megabits per second.

AirPort wasn't disruptive; it was convenient. It was a clever idea that appeared to pose no threat at all to the communications-infrastructure players. What threat could be posed by a 100-meter network? But there is power, real power, in WiFi's unlicensed nature. As long as users didn't break the FCC rules on Effective Radiated Power (four watts max), they could do anything, and this was very attractive to experimenters, who came up with all sorts of WiFi-based business ideas, two of which eventually stuck. The first great idea was the so-called hot spot, a WiFi access point connected to the Internet that allowed those nearby to make a wireless connection to read their e-mail or look at dirty pictures while they drank coffee. Devised in the Internet's heyday, the not-very-well-thought-out idea was that people who could go online would buy more coffee and pay some sort of fee for access.

Companies you have never heard of are creating networks of hot spots in restaurants, in airports, on trains.

There were no great fortunes made in the hot-spot business at the time, primarily because hot spots aren't cheap to install and run, and there was no regional or national infrastructure to allow potential users to be linked to the Internet when they weren't drinking coffee.

Only recently, with equipment getting cheaper and millions of people set up for WiFi, are national hot-spot networks starting to appear. Companies you have probably never heard of, like Boingo wireless, Joltage, hereUare Communications, iPass, WiFi Metro, and Way Port, are creating networks of thousands of hot spots in restaurants, in airports, even on commuter trains and airliners. They are aggregating hot spots and creating value where it did not exist before.

Hot-spot aggregation relies on Metcalfe's Law (coined by Bob Metcalfe, inventor of Ethernet networking and a former boss of mine), which says the value of a network increases with the square of the number of its users. So a network with 1,000 users is worth 100 times as much as a network with only 100 users. And a network with 1 million users is worth a million times as much as a network with 1,000 users. The bigger the network, the better. This points to a major shakeout coming. Soon the aggregators will themselves be aggregated, and two larger companies will probably emerge. And they'll have the privilege of competing with Intel, IBM, AT&T, Apax Partners, and 3i -- which recently announced Cometa Networks, a joint venture to "install" (or "acquire") 20,000 hot spots in the top 50 U.S. metro areas.

It is probably too late to make a fortune in the WiFi aggregation business, but there is plenty of opportunity in the hot-spot business, since most buildings and public places are not yet networked. Put one in place, and wait for the networks to come calling. With aggregation and super-aggregation pending, having a flower shop in Grand Central Terminal might prove more valuable for its hot-spot potential than its flowers.

Hot-spot value is a matter of location, location, location, but much also depends on the type of technology you use. This brings us to the second great WiFi innovation -- high-gain antennas. If you look at an Apple AirPort base station, it appears to have no antenna at all. Even the Linksys WAP11 wireless access point, which is among the least expensive and most popular WiFi devices, has just two little rubber antennas, each about four inches tall. These are low-gain antennas, which explains the limited range of such devices. But it is possible to put much larger high-gain antennas on wireless access points, greatly extending their range.

And I mean greatly. My own record for longest WiFi connection is approximately seven miles, which isn't bad for a technology normally thought to run out of gas at 300 feet. The absolute record to my knowledge is (so far) 72 miles! And this is all perfectly legal and within FCC specifications. So WiFi can be used for long-distance connections too. For more than a year, I got my Internet service that way here in California wine country, and it never failed.

Long links have value by themselves, but that value can be greatly enhanced for business if high-gain antennas are used to serve hot spots. Here the definition of "location, location, location" changes. Hot spots typically require a high-speed Internet connection to the location and the presence of a WiFi base station, both of which need the permission of whoever controls the space. Let's say it's Starbucks. And let's figure Starbucks probably wants to be paid.

High-gain antennas can change that situation in some cases, according to Dan Arra, president and CEO of Gatespeed Broadband, a wireless Internet service provider in San Jose, Calif. "We use our existing Wireless Points of Presence to blast a group of stores, coffeehouses, restaurants from atop a building," says Arra. "We don't do this everywhere, only where we already have bandwidth and roof rights. There is no extra cost to us except the access point [base station]. I set one up this morning. I put an access point with an 18dBi directional antenna on top of our downtown San Jose [point of presence] and pointed it at the Starbucks, Rock 'N Tacos, Spiedo restaurant, and Campbell Cigar shop below. It works great. I got 1.2 megabits per second inside these places with my WiFi card. I didn't have to ask Starbucks or offer to pay them anything!"

Since there is no way a business can legally exclude radio waves from its premises, Gatespeed's hot spots are much less expensive to build, especially since several such hot spots can share a single Internet connection, where previously they each would have required a connection. When the super-aggregators come calling, Gatespeed won't have just a few restaurants to offer, it will have most of downtown San Jose.

Say you run a business with offices near the top of a tall building in almost any city. Becoming a wireless ISP is as simple as suction-cupping a high-gain antenna inside one of your windows and connecting it to your high-speed Internet connection. Customers use their own high-gain antennas, mounting them outside their buildings or suction-cupping them inside their own windows facing your location. It doesn't require much precision.

But what if you want wireless Internet service and your window is facing the wrong direction? That's where yet another technology comes into play -- a technology that can ratchet up Metcalfe's Law yet another notch. Called HomePlug, it uses the electric wiring of your home or building to share data at speeds up to 14 megabits per second over distances up to 1,000 feet. HomePlug effectively puts a network connection at every electrical outlet, since that's the distance between electric outlets set by the building code. The business opportunities here are obvious. Let an apartment building share a single Internet connection, perhaps that fixed wireless Internet connection from the other side of the building. It could be the landlord who sets up this business, adding the Internet cost onto the rent. Or it could be the tenant in apartment 3-C who has that high-gain antenna on his window or who has a business DSL (you can't legally share residential DSLs).

As long as people are on the same electrical transformer, they can communicate, so you can serve half your neighborhood. Recruit your friends, and serve half of every neighborhood. This is high tech but low cost, and most people don't know about it because the first version of HomePlug was limited to 16 devices. But the next version, shipping now, is good for up to 128 nodes. And the version after that, which will appear in 2005, supports HDTV! Suddenly, and entirely because these products are minimally regulated, the phone company and cable company are both in jeopardy. Now that's disruptive technology.

Contributing editor Robert X. Cringely is a writer, broadcaster, and entrepreneur specializing in high technology. Contact him at cringely@inc.com.

Please E-mail your comments to editors@inc.com.