OK - is Bluetooth really doomed?
UWB may not go USB's way...
Bluetooth, the wireless communication standard built into mobile phones and other devices, has weathered plenty of storms in the past. This week, it's been suggested that it has killed itself off by its choice of a faster technology. The group's marketing boss thinks otherwise.
Just when things are going well...
Things should be looking good for Bluetooth. After years of seeming to be more of a European idea, the technology has finally broken into America, along with decent cellphones. A study commissioned by the Bluetooth Sig found that half of American consumers are aware of Bluetooth, compared with around 60 percent for Wi-Fi. Here in the UK, 88 percent of us know what Bluetooth is for.
Just when everything should be going fine, up pops Martin Rofheart, of UWB chip-maker Freescale, to tell us that Bluetooth has doomed itself with this decision.
What's that again?
Rofheart's statement isn't that big a surprise. There are two major rival versions of ultrawideband, and the Bluetooth Special Interest Group chose the Intel-backed Wimedia flavour over Freescale's DS-UWB.
We warned a year ago, when Bluetooth first said it would be using UWB for its next, faster version, that the choice of an actual technology would be an incendiary issue. Turns out we were right.
Freescale's DS-UWB got sidelined by the USB Implementors' Forum, and only achieved a stand-off in the IEEE, which gave up on the idea of a UWB standard. Now Freescale reckons that getting first to market will be enough to make it the de facto upgrade for USB, and to knock out Bluetooth by doing the same job sooner.
Why go for a narrower band?
The Bluetooth SIG has chosen Wimedia, and has specified a narrower frequency range than the mask approved by the US regulator, which will be used for wireless USB products.
In so doing, says Rofheart, the Bluetooth SIG has condemned itself to being late, being more expensive and performing worse than wireless USB. Wireless USB will be the de facto wireless personal area network, says Rofheart, and because of the big overlap between the jobs they do, wireless USB will sweep away Bluetooth.
Anders Edlund, marketing director of the Bluetooth SIG, isn't impressed with all this:
"Our reasoning behind this decision has been that Bluetooth needs to continue being a global standard," he says. "We are working with regulators across the globe and it seems that the above 6 Ghz requirement will be a feasible way forward."
The FCC mask is the only one approved for use at the moment, but Edlund reckons it won't get adopted worldwide without big changes. "Potentially this may happen, but it seems this is not the route the regulators are taking at the moment," he says. "Lobbying obviously plays a role here and the mobile operators have a very strong say in this process which is why I wouldn't place my bet on [universal adoption of the FCC mask]." It is the mobile operators that have largely influenced Bluetooth's choice of frequencies above 6Ghz, he says.
But surely the US will drive it?
We suggested that the US mask will carry the day, especially when there are mass market products that meet it. "I agree that this might be a possibility," concedes Edlund, acknowledging that this is exactly what happened with Wi-Fi and the European Hiperlan standards. "It is the US that drives computing and since Wi-Fi/Hiperlan was mainly about computing it was never hard to guess who would win that battle. For technologies spanning many other usages it is different and Bluetooth has so far been driven more form Europe and a mix of industries."
It looks as if Europe and Asia are going a mask that limits UWB to frequencies above 6Ghz spectrum, he says. He admits this might change, but with this uncertainty, "in our opinion it would be bad for the industry as well for the consumers to launch devices sooner that will only be allowed to work on the US market."
So will UWB-Bluetooth be late? "We have obviously also worked with major silicon developers and from an implementation point of view we believe that in the next couple of years cost-efficient chipsets will be available," he says. "Since our target is to have the spec approved by end of 2007 and expect products in 2008, that coincides reasonably well with the time frame when both regulatory approval is expected as well as silicon in volumes."
Bluetooth versus wireless USB?
In any case, he doesn't believe wireless USB will beat Bluetooth in providing a personal area network. "There are three and possibly even four different WUSB standards emerging," he says. "There is also a rather big paradigm shift from USB to wireless USB with plenty of nasty obstacles to solve."
Wireless USB will have to develop robust ways to handle issues that Bluetooth has already got down pat, such as pairing, power consumption and user education. Wireless USB also looks like being limited by a lack of device-to-device connectivity since a host is required.
There will be interoperability problems between the different wireless USB technologies, he says, and maybe between the same solution, if any of them goes world wide in a year or two. If there's a different frequency mask, would a wireless USB device designed for the US work with one designed for Europe?
"With Bluetooth most of the issues are sorted out and many users understand the Bluetooth technology," he says. "I personally think the interesting question is how USB can move from cable to wireless without losing all the strengths it has. If you think about the typical USB scenarios I personally see some problems in this.
And how about bandwidth?
Edlund denies Rofheart's prediction that the narrow frequency mask will cut UWB-Bluetooth's speed to 10Mbit/s at 10m. "We expect performance up to around 500 Mbit/s over a metre or so, using the frequencies above 6 Ghz," he says. "We are expecting to have about 100 Mbits/s at typical Bluetooth range (10m) which we believe is sufficient for most Bluetooth applications.
"We see no issues with performance," he goes on, "even though the band is narrower. The other lower band alternative would in any case need to mask out large chunks of the band so in reality there might actually be more available bandwidth for the higher band."
Watch this space
The frustrating thing is that Bluetooth's stately progress means that, even if Cable-Free USB devices do appear this summer, there will be no way to know how they will stack up against wireless Bluetooth, until that arrives.
If we really are all using Cable-Free USB in 2008, then it is game over for Bluetooth. But Edlund's picks some sizable holes in Rofheart's argument. It's going to be interesting to see what happens.