The Future of Mobile Connectivity

How this market will shake out is anyone's guess

Mobile devices are an interesting social and business phenomenon. Over the past 10 years, businesspeople have come to expect to be constantly in contact with the office. For most of us, a cell phone makes this contact possible. Although cellular communications are not as ubiquitous in the United States as in many European countries, it's hard to imagine a US businessperson without access to some form of cellular communication. Cell phones have become a communication appliance that's just as common as the wired telephone in your home or office.

The PDA Paradox
PDAs are trying to achieve that same level of universality but tread a difficult path that contains two major obstacles—one social, the other technical. The social obstacle is an interesting paradox: Few people who have tried PDAs are ambivalent about them. Users either love PDAs or hate them. (Try getting that worked up about a telephone.) Plenty of people buy cell phones for only occasional or emergency use, which has helped create universal acceptance. But becoming a PDA user is rarely spontaneous—you either reorder your life around your PDA or you don't use it at all. PDAs don't lend themselves to casual use.

The technical obstacle is one that affects every mobile device, not just the PDA: lack of connectivity. A mobile device that can stay in touch with the rest of the connected world has a significantly higher business value than one that is an island unto itself. For the most part, the only connectivity most people experience with their mobile devices is during the times when they are connected to another machine or to their office network. The one notable exception to this rule is Research In Motion's (RIM's) BlackBerry wireless device. You can spot a BlackBerry user from across the room by the typical posture: head down, shoulders slouched, eyes squinting, thumbs banging away on the tiny keyboard. With just under 300,000 users, the BlackBerry is, by orders of magnitude, the most successful connected device. But it's fundamentally a PDA, which means it suffers from the social obstacle I just described. There are no casual BlackBerry users; all I've met would be lost without their BlackBerry at hand.

Duking It Out in the Marketplace
Other connectivity options for mobile devices have yet to make their mark. Metricom's Ricochet wireless network service showed promise for mobile users, with 19.2Kbps connectivity that Metricom upgraded to 128Kbps in some areas before the company went under. This service let you take any device with a PC Card slot on the road in many major metropolitan areas and provided Internet connectivity at decent speeds, especially with the 128Kbps upgrade. The problem was that not enough people were clamoring for the extra mobility Ricochet gave them at the price Ricochet wanted to charge. The entire Ricochet network was sold as a block (and is still functional—the New York City segment was turned back on to provide connectivity to workers during the aftermath of the September 11 terrorist attacks), and the current owner, Aerie Networks, has made clear its intention to reactivate the network more cost-effectively.

The current darling of the mobile connectivity world, the Bluetooth wireless specification, has yet to make a significant impact, nor is it found in any significant devices. As I write this, the Bluetooth SIG Web site lists more than 500 available Bluetooth products, but I've yet to encounter one in the real world. (That statement isn't entirely true: A coworker's IBM notebook was delivered with Bluetooth built in, the only one in our order of a half-dozen IBM notebooks to be so equipped.) I'm not encouraged when I received a press release touting that the latest Bluetooth advance will allow shoppers to receive coupons that a cash register beams directly to their PDA at their favorite retail merchant. Such applications won't bring about widespread business adoption.

Today, the only wireless connectivity method that I consider an unqualified success is Wi-Fi, the 802.11b wireless standard. 802.11b is fast (11Mbps), cheap, and most important—it works. The specification isn't perfect, but it's in wide use, and as usage increases, the changes that are necessary to make 802.11b available everywhere will fall in place. Computer groups in San Francisco are placing public wireless access points throughout the city, allowing roaming users to stay connected when they're not tethered to their office. Some companies have announced plans to make 802.11b networks available in locations with heavy business traffic, such as airports, and Amtrak has announced 802.11b Internet connectivity availability on some of its trains.

But even with this wide network availability, 802.11b has serious problems that need solutions. For example, I use mixed 802.11b and wired networks in both my office and my home. I can plug network cable in at either location, get an IP address from the DHCP server, and be up and running in seconds. With my 802.11b networks, the situation isn't so simple. At my office, we run a high-security network, with a lot of encryption and authentication. To use the same configuration at my home, I would have to add substantial complexity to my simple wireless network small office/home office (SOHO) environment. Because my home network needs to be accessible to nontechnical users and I don't want to spend my rare hours at home doing tech support, I got into the habit of reconfiguring the wireless card in my notebook for whatever network I happened to be using. I found it simpler to just buy another wireless networking card (of a different brand) and configure that card for my home network. Now, switching from home to office wireless configurations is just a matter of switching PC Card NICs. But this solution isn't practical for most users.

The Future Is Wide Open
Perhaps mobile connectivity is becoming a chicken-or-egg proposition. Does someone need to build the infrastructure and wait for the clients to appear? Or do we build the client base before building the infrastructure? And even if the infrastructure and the clients appear, who's responsible for the authentication? Where is security handled? How do you move between public, private, and semi-private networks?

Corporate systems administrators aren't likely to trust users to make the right decisions, particularly regarding the security of their private networks (I already hear regularly from systems administrators who lament that their corporate policy lets users take notebooks out of the office). Consider the ambiguity of this situation a heads-up for network administrators. The only way to have an impact on the future of mobile connectivity is to vote with your dollars. The trick is to spend those dollars wisely and at the right time.