Update - July 16 - Alan Reiter in his July 16th entry in Reiter's Wireless Data Web Log weighs in with a thoughtful analysis of the prospects and likely perils of Project Rainbow. Infoworld weighs in with some additional comments.
In the July 15, 2002 issue of the New York Times, John Markoff reveals the existence of "Project Rainbow", a project that considers creating a national 802.11b network for portable and mobile use. Partners in the venture include Intel, IBM, AT&T Wireless, Verizon, Cingular and unspecified others. Markoff states:
The discussions, which are code-named Project Rainbow and have been going on for the last eight months, envision a nationwide service that would provide on-the-go professionals and other Web surfers a unified way to reach the Internet from a wide range of "hot spots" like airports and other public places. It is not intended to supply broadband connections to customers' homes, an executive involved in the discussions said
From the cast of players, there's a little in this for everyone. IBM ties the network together. AT&T Wireless, Cingular, and Verizon provide the sites. Intel sells more end-user devices. And that's where this begins to fall down. To really make this work will require a bold pricing model and ubiquity. Alan Reiter, quoted in the article, concurs:
"I think it could jump-start the industry if all the components are in place," said Alan Reiter, publisher of Wireless Internet and Mobile Computing, a wireless data newsletter and consulting firm in Chevy Chase, Md. "That has been the problem with wireless: everything has to work well or consumers will reject it. You need the right pricing, the right devices and right locations."
I have no fear about the devices; that's being rapidly taken care of by the rapid emergence of 802.11b built-in to PDAs and 802.11b SD cards (which have emerged as the preferred form factor for PDA expansion cards.)
It's hard to imagine this cast of giant companies being willing to work together harmoniously enough, long enough to make this a reality. There will be way too much temptation to focus on the needs and fears of one or more companies. What happens when users begin making VOIP "telephone calls" from Project Rainbow hot spots? If they're enlightened, they'll realize this is incremental revenue.
If Project Rainbow somehow survives its formative stages, then it becomes a technical problem - how to scale so many 802.11b APs to provide ubiquitous coverage? There are many, with perfect validity, that argue that you shouldn't try - that 802.11b isn't the right technology for this purpose. It's notable that Metricom's (now Aerie Networks) Ricochet system was widely perceived as a "good enough" solution to the scaling problem that Project Rainbow will face, and used a lower data rate (128 - 256 Kbps) and completely different technology (FHSS) and spectrum (902-928 MHz band) to achieve it. Other "waiting in the wings... just add spectrum and money" technologies are ArrayComm's iBurst system now being deployed in Australia, IPWireless' system using MMDS spectrum, using PCS channels as data-only as Monet Mobile Networks is doing, Flarion Technologies... and others.
But, it sounds like Intel, from the tone of the article is the formative influence of the group, understands clearly that the end-goal is to address the enormous opportunity that will soon emerge when PDAs are equipped with 802.11b capability. It's 802.11b that, for better or worse, must be accomodated. I liken this problem to the continued existence of serial and parallel ports on the average PC - there are much, much better ways to communicate between peripherals... but it took the better part of two decades for a next-generation communications link - USB, to achieve ubiquity. Ubiquity trumps technical elegance, and Intel, having driven USB into ubiquity, understands that better than most.
What would it take for Project Rainbow to become reality? That's an easy call - money. Lots of patient money. Billions... think Microsoft, with its cash hoard of $35B or so.
7:06:55 AM 
|
