The IEEE's latest work may provide Bluetooth's death knell, but not for four years: I write in an upcoming InfoWorld print edition (this is the advance online version) about the work being done in the Wireless Personal Area Neworking working group (802.15)'s high bit rate task group at the IEEE. This group is focused on fast, easy, and low power for streaming media and other big files.
The leading contender for the physical layer at 802.15.3a is ultrawideband. One proposal has garnered more votes, but some issues have to be resolved for it to be adopted. If it can't pass, another UWB proposal probably will, but the winner won't show up in silicon for a year or so and won't be widely available and cheap until 2006 or 2007.
What's the joy with UWB? Very high speeds over very short distances. Imagine walking in with your laptop and even as you get near your office, a file transfer starts that, in a minute or so, moves hundreds of megabytes between them. Imagine plopping a hard drive next to your computer and seeing it mount. Imagine wheeling a shared scanner into your office and just starting to work. Imagine having several monitors, none of them connected, all of them providing extremely high-quality displays -- and your computer is in a closet next door.
None of this is even remotely science fiction; it's the logical evolution of wireless PANs in the context of personal and enterprise computing.
Bluetooth is not an inherently bad spec, but it has two key failings, which were limitations of the time it was developed: the first is speed. 1 Mbps turns out to not be a USB replacement except for very low-speed peripherals like keyboards. It's even frustrating for synchronization. That was a feature of the engineering possible at the time. (802.15.3, not .3a, provides a kind of Bluetooth migration using spread spectrum in 2.4 GHz but at 11 to 55 Mbps; it was just finalized and it on its way to approval, but .3a stole its thunder.)
The other failure was the difficulty of use. Recently, some Bluetooth SIG marketing folks have talked about getting down the time to hook two Bluetooth devices together to five minutes. I'm an experienced Bluetoother and it still takes me a bunch of monkeying around.
The 802.15.3 task group has simplified connections to avoid these problems, and .3a kicks the speed up to 110 to 480 Mbps. These two changes alone will make Bluetooth potentially superfluous when .3a hits because .3a also has good battery life.
Another group has worked on extremely good battery life: 802.15.4, which is already being marketed as ZigBee with no products in sight. ZigBee wants six months to two years of usage out of, say, a couple of AA batteries. Because its intended to replace infrared controls and other devices that use wires for very very little data (like alarm circuits), it might wind up using UWB running at just a few tens of kilobits per second.
The clear outcome of 802.15 in general is massive success: they've taken a niche part of the industry, gotten huge semiconductor firms involved, and potentially moved us into the wireless office of tomorrow -- we just have to wait for it to arrive.