Home > Mobile and Wireless

Computerworld - A future built around pervasive wireless communication ranks high on IBM's list of the most important trends in "on-demand" computing, right up there with autonomic, or self-healing, networks, ever-speedier processors and modular hardware and software designs. "Wireless is very, very big for IBM as a growth area, but it's much more than just integrating Wi-Fi radios into ThinkPad [notebooks]," said Brian Connors, chief technology officer of IBM's Personal Computing Division. "It's very strategic for us in productivity and ease of use, as an enabler for access to anywhere-anytime information."
With global responsibility for technology trends, directions and investments affecting the PC, Connors has a hand in everything from business development, standards-tracking and partner relationships to research directions. He spoke yesterday with Computerworld editor in chief Maryfran Johnson about the proliferation of Wi-Fi "hot spots," corporate stumbling blocks to widespread wireless use and IBM's research directions.

Brian Connors, CTO of IBM's Personal Computing Division

What is IBM's strategy for Wi-Fi? Wi-Fi is happening -- it's not a matter of if or when -- and it's coming at us from a lot of grass roots. It's the customer looking at a three-year life cycle, and they know their users might be in hot spots outside their office. From the IBM perspective, it's huge. There are tremendous integration services to be built around wireless. From the PC business perspective, we see wireless as a key enabler for access. It's very strategic for us in productivity and ease of use. But we're enabling the platform, we're not engaged in those business models where companies like T-Mobile or Wayport are creating value as a service offering.
Do you see a clear technology winner when it comes to providing wireless service? The user model is going to define all this, not the technology model. If wireless is fast, then give me the fastest network. Wi-Fi is high-speed, standardized around the world pretty much, unregulated short-distance frequencies. It's an extension of Ethernet, basically. Wide-area networks are regulated, not standard around the world, and WANs have a cell phone billing model in place. More and more you're going to see things like wireless WANs, like what AT&T and Sprint and other carriers are providing now. You don't have to discover the service through the wireless network, it will be persistent and available, and quality of service relative to bandwidth will keep improving.
Wouldn't the Holy Grail for Wi-Fi be entire cities engulfed by a Wi-Fi cloud, making it feasible to provide applications like low-priced voice over IP or videoconferencing on a multimedia device? You have to ask, "What's the business model? Who gets paid?" High-speed wireless is important, whether it's Wi-Fi or WAN. Quality of service is what will define it. You want anywhere-access to data and information, made available at a cost-efficient rate. I'm very interested in the billing models for the telcos and the wireless cell providers. There's a good opportunity for them as they partner. They can see the disruption of Wi-Fi, which is why they must embrace it. I think they're starting to.
But some analysts say all the hoopla about Wi-Fi hot spots in locations like Starbucks or McDonald's boils down to a clever marketing gimmick, not a long-term trend in wireless access. What do you think? Well, I very rarely take my ThinkPad into a Starbucks, but if you look around college campuses, it's very popular in all kinds of common areas. Hot spots as you know them are really just an area of concentration of users -- natural places to get your e-mail, like business hotels or airports. Secondary places are the congregation points like Starbucks, bookstores or campuses. The third area, where you see wireless going more vertical, is in dealing with consumers at touch points in retail stores, for example. Those are places you'll see more and more wireless. In business and in the enterprise, Wi-Fi is going to be there. Once you use it, you can't live without it.
From an IT perspective, what are the biggest stumbling blocks to widespread use of wireless? Wireless communication is becoming pervasive, but I say pervasive because it's not guaranteed. There are a lot of potential inhibitors, including bandwidth, coverage, usability and security issues. IT has to have confidence they're not exposing the network. A lot of the benefits of wireless are around personal productivity. The lines of business want to deploy it, but the IT organization has to ensure a secure environment and protect corporate assets. I do think wireless is the future, but I don't think we'll see high-speed connections available just any place soon. Standards are important to watch, and looking at what's emerging from the universities and usage patterns of consumers.
We've written a lot about the problems posed by unauthorized or rogue wireless access points inside companies. How do you deal with that internally at IBM? Three years ago, in the ThinkPad team, we started using [an internal product] called Wireless Security Advisor [WSA], which gives you the ability to walk around and conduct a wireless audit. Our Tivoli division and our services organization also use WSA to secure the enterprise in batch jobs. Now, we've taken that [technology] a step further with the creation of Distributed Wireless Security Auditor, which is using ThinkPads to autonomically sniff and monitor the environment. We use it internally and in consulting operations, but it's not in general availability now. We're looking at that seriously, making it generally available.
Does IBM plan to produce a low-cost hot-spot-in-a-box product like Toshiba has done? Will you get into handhelds or tablet PCs at some point? No. We have no plans there. With tablet PCs, we see more people piloting and evaluating, but it's in the very early adopter phase.
Last year at Comdex, you talked about an "Instant Connect" feature coming to ThinkPads and other PCs that could seamlessly detect a connection -- wireless or wired -- and configure itself with very little user intervention. How's that coming? We're planning for the first half of 2004 for that. Instant Connect is meant to tunnel its way out and aggressively figure out the problem, say if you're at Starbucks and getting denial messages or problems connecting. You'll also start seeing these capabilities evolving in our Access Connection product.
What else should customers be looking for from your division? More innovative designs and longer battery life. When we look at batteries, one of the things we worry about is deteriorating battery life when you get to Wi-Fi 802.11a in the 5-GHz range. It's a power hog. So in battery and system design, we've done a lot to maximize the battery life. We're researching technologies now on how we can sense proximity of ThinkPads to access points and how we can shut down the power output to scale to the relative distance to an access point. We don't know what it'll yield yet, but we're researching that area.