|E-mail Wes Simonds|
It seems IBM's interested in doing a nationwide wireless local area network (LAN).
"Some customers just want it on their campus. Others are saying, 'Wouldn't it be great if there was some national network?' It seems so obvious that it should be this way," said Dan Papes, vice president of IBM Global Services' wireless services division.
Now, this is a very logical idea, and one which is paired to a demand and infrastructure which are growing nicely in parallel. The U.S. already has an increasingly diverse assortment of wireless LAN "hotspots" here, there, and everywhere (as Paul McCartney once said under different circumstances), but what road warriors really want is ubiquity -- truly universal access regardless of location within the city. A city, in other words, should be a hotspot -- a single, integrated, robust, scalable hotspot.
In the case of wireless data, even more than wireless voice, you're going to want your nines. You need the Internet, you need your virtual private network, you need them everywhere you go, and you need them to work without fail.
So I've got no beef with IBM's ambitions. They seem very sensible. One question I find myself asking, though, is what IBM has to bring to the table.
Does it own the protocols? No. The chips? No (those come almost exclusively from Intersil, Lucent, and a few scattered minor players like Atheros and Radiata). The management and quality-of-service solutions? Nope. How about the existing nationwide infrastructure that can be leveraged to deploy a nationwide wireless LAN of the type described? Not to my knowledge. It's got the brand name, of course; it's always got the brand name.
There's also the question of whether IBM's going to find the profits smaller than the price. As you might expect if you read my last column, IBM is not exactly an innovator in this regard. The notion of a high-bandwidth wireless solution that plays nicely over entire metropolitan markets has been explored by other people, and the results have been mixed at best.
Most notably, consider Metricom, which in Ricochet provided a surprisingly high-performance wireless solution in a surprising number of markets. I never heard a single negative thing about Ricochet from anyone, except, finally, this: that it dried up. The bottom line was unfortunately that Metricom wasn't prepared to deal with the costs of expansion and the oscillations of demand. Now, this might not be a problem for Big Blue, which has very deep pockets indeed -- pockets deep enough to hide leviathan sea monsters like OS/2, for instance -- but one still wonders if the money might best be spent elsewhere.
Yet another factor for IBM to consider, if it hasn't already, is the possibility it could be starved out by grassroots wireless LANs. Have you heard about these?
The idea is that Samaritanesque individuals, working in cooperation and using common protocols such as 802.11b, open up their wireless LAN access points in densely populated urban environments, thus providing low (or no) cost wireless Internet access to entire communities. Such networks have been informally popping up in roughly the same areas as commercial wireless LAN hotspots -- along the coasts, for the most part. Seattle, San Francisco and New York are all leaders in the field.
The obvious technical objection to the idea is simply that it's hard to picture the coverage being very complete; the normal range of wireless LAN access is only several hundred feet. However, many of these folks are using hot-rod technology designed to improve that radius. The expansion of service is dramatic in some cases; I've heard of cases such as one in Hawaii that involve a wireless LAN more than 20 miles wide.
The concept of shared bandwidth is, to some degree at least, bound to threaten commercial nationwide wireless LANs. It has the charm of universal demand coupled with the David versus Goliath selling point commonly associated with open-source software. It, in short, stirs buzzwords into something greater than the sum of their parts.
What do I mean by that? Well, here's a taste sample on a little wooden spoon: peer-to-peer wireless. It's Reese's peanut-butter technology: two great tastes that taste great together, and it's growing as a trend, not fading. Consider that in direct opposition to Napster, this is a form of peer-to-peer that uniquely doesn't consume bandwidth; it provides it! It's easy to see why it's gaining strength.
Of course, it's quite possible that the enterprise market IBM's chasing is so different there's no real issue of competition. Ad hoc grassroots networks are intrinsically less appealing to businesses than consumers. They're unpredictable, and in the event of problems, very probably unanswerable. And it's not hard to imagine major Internet service providers/carriers becoming increasingly uncomfortable with them and moving against them legally and technologically, much as the music industry did successfully with Napster. So the cash-flush players like IBM are probably less than intimidated.
Still, I find the joint development of commercial and grassroots wireless LANs on a nationwide basis a fascinating horse race, and I'll be curious to see the outcome.