Online Extra: Where Wi-Fi Goes from Here

Intel's Centrino guru Jim Johnson says quick, no-hassle Internet access from just about anywhere isn't that far away

Wi-Fi -- short for wireless fidelity, a technology allowing anyone with a laptop and a wireless card to get high-speed Internet access at places like cafes and airports -- has been one of technology's few bright growth spots. In 2003, about $1.67 billion will be spent on Wi-Fi equipment worldwide, according to Allied Business Intelligence, in Oyster Bay, N.Y.

Wireless devotees can thank Intel (INTC ), the world's largest semiconductor company. In October, it pledged to invest $150 million over several years into Wi-Fi startups. In December, it joined with AT&T (T ) and IBM (IBM ) to create Cometa Networks, which is setting up Wi-Fi "hot spots" (places that allow for wireless access) in 50 U.S. cities. And in March, Intel unveiled its Centrino brand of chips for laptop computers. The new chips have Wi-Fi compatibility built in. In fact, according to a recent forecast from market consultancy Cahners In-Stat, 90% of laptops sold in 2005 will have built-in wireless-access capabilities, up from 5.7% in 2002.

Intel's efforts are one of the forces behind Wi-Fi's recent boom. Although Wi-Fi was standardized in 1996, only a few years before the Internet boom started, it's still in its infancy, says Jim Johnson, general manager for the wireless networking group at Intel and the prime mover behind the Centrino chips.

Intel's researchers are already working on new generations of wireless technologies, enabling faster, more secure high-speed access. Some of these networks will allow access from miles away -- a huge improvement over today's 300-foot radius from the access point. Johnson talked to BusinessWeek Online reporter Olga Kharif on Apr. 14 about Intel's wireless vision and its research projects. Following are edited excerpts from their conversation:

Q: What are Wi-Fi's drawbacks?


One of the concerns I run into the most talking to enterprises is security. Early on, hackers could break into the security keys used. So most enterprises treated their wireless network like their remote-access network. They put authentication servers in place that use virtual private networks (VPNs). That's what we've done to make sure our wireless network is robust.

However, to keep adding more users, we need better technology -- better authentication, encryption, and better encrypted transport of data. We are [one of the companies] working on a new technology, likely to come out in the first half of 2004, which should address these problems. In the meantime, a bunch of us -- Intel, Microsoft (MSFT ), Cisco (CSCO ) -- have developed an interim security solution called the Wi-Fi Protected Access, which will be shipping in the middle of this year. Then, enterprises could either go the VPN route or use these new technologies to increase security.

Q: Analysts also say for Wi-Fi to realize its full potential, its reach has to be extended. What are you experimenting with in this area?


We're running a pilot out of this building: We are getting Wi-Fi to a home 10 miles away, running at 2 to 3 megabits per second, and a home 20 miles away at 1 megabit per second [using directional antennas mounted on a roof at Intel's Jones Farm Campus in Hillsboro, Ore.]. As the distance from the access point is increased, the speed drops off.

Q: What do you think Wi-Fi would look like in five years?


I envision this five-minute experience: If you're in a city, you'll know which coffee shop or hotel has a high-speed connection within five minutes of where you are. And if you don't need a high-speed connection, you'll use traditional wireless carriers' networks from the same device.

So you'll always be able to make a connection, but you also could get a higher-performing one. In a village or at a suburb, you'll have the same five-minute rule: Places with high-speed access might be a five-minute drive away.

Q: What are you doing to make this vision come true?


What our labs are focused on is, first, developing low-cost silicon technologies so you can embed radios in multiple locations so users can find the networks. Second, we're working to improve ease of use and price and performance of the devices and the networks.

You wouldn't use your mobile phone without it working just about wherever you are -- and this applies to Wi-Fi also. So we'll eventually integrate all of the different wireless technologies, including those used by wireless carriers and Wi-Fi, onto chips that will go into the laptops, so you could easily access any network, wherever you are.

Q: What else can Intel do to make Wi-Fi a success?


Right now, you open a notebook with Centrino, and you see all the access points there are around you. I was at an Austin airport, and they had seven or eight access points for me to choose from, and each one had its own account. So I had to find the one I had an account with. But I'd much rather be able to have one account and have all service providers figure out how they're going to get their share of the fees I pay. I want to know that, as long as there's a Wi-Fi hot spot near me, I can use it. That's still quite a ways off, but it's absolutely required.

Service providers have to figure out how to make this work, but we're doing something that I know is going to help. Our Centrino products, enabling mobility, will speed up Wi-Fi adoption. And with millions of Centrino notebooks out, carriers will be motivated to resolve these issues. Plus, we make investments like Cometa to try to spur roaming agreements.

Q: What are you doing outside of Wi-Fi?


We have engineers working on a variant of Wi-Fi, called WiMAX, which will have a 31-mile linear service range -- a huge improvement over Wi-Fi's 300-foot radius. It will also offer much higher speeds, of up to 70 megabits per second, vs. 11 for Wi-Fi. I think there will be lots of variants of Wi-Fi.

We're kicking off a high-throughput group, which will work to increase the networks' speed. But our mantra is getting significant performance improvements for similar prices and with backward compatibility, so you could have a new PDA (personal digital assistant), for instance, and still use the old technology.

Q: Where would WiMAX be used?


Down the road, different standards will win in different applications, so we're going to invest in multiple paths. Wi-Fi looks obvious now, but there were other competing technologies before. We invested in all of them. As soon as we saw Wi-Fi was winning, our resources shifted. But I think Wi-Fi is going to win long term, too, [in laptops and PDAs].

WiMAX might be used [to connect devices] inside the home or for connections between Wi-Fi access points. We'll find the best technology for the best application, and then we can combine different technologies.

Q: What's in all this for Intel?


One of the main reasons why folks buy computers is to communicate with other people. With a desktop, connecting involves going through a phone line or a broadband connection. With a notebook, you can be mobile. We believe we can't deliver value without communicating. Our mission is to bring communications to other computing devices.

Q: Do you expect wireless broadband to drive computer replacements?


Given today's economic environment and capital expenditures, what's the reason for companies to spend money? It's to get more out of your fixed resources. And I absolutely believe these technologies make people more productive. We've installed Wi-Fi into Portland's Jesuit High School. A teacher can now do two jobs at the same time: for instance, monitor students in the hall during recess and write lesson plans.

We believe that if we change the way people use computers, they'll use computers more -- and more people will want to use computers. That's a big benefit for Intel.

Edited by Patricia O'Connell

    Before it's here, it's on the Bloomberg Terminal.