Bluetooth's low-power variant could be everywhere starting next year, and other low-power networking options had better watch out, say the new standard's backers.
Last year, a Nokia technology called Wibree was adopted as a low-power variant of Bluetooth, by the SIG (special interest group) which directs the standard. The technology was demonstrated earlier this month under the "engineering name" ultra-low power (ULP) Bluetooth.
The technology is due for a new "marketing name", we hear. It's apparently going to become "Bluetooth low energy", and start to contend with rivals. It will be a tough contest, and the winner will be the one that is built into most devices, costs least to implement, and use least power - so devices' batteries will last longer.
Technology comparison One technology that should look to its laurels is ZigBee, the low energy wireless mesh network that has been on its way to household appliances for the last five years. Bob Heile, chair of the ZigBee Alliance claims not to be worried by low-power Bluetooth: "I don't put them on the radar screen," he told Techworld. "It's just Bluetooth with frequency hopping turned off. It's not a networking solution, as it allows no more than eight devices."
Maybe he should have them on the radar, because he's wrong, according to Robin Heydon of wireless silicon maker CSR: "Other people don't understand the details of the specification," he says. Going to low power gave the designers a clean sheet of paper, he says, so low-energy Bluetooth can share antennas with regular Bluetooth, but have a completely different MAC. Regular Bluetooth has a limited number of connections, but its low-power sibling doesn't says Heydon.
"The number of devices is unlimited. The practical limitation depends on how much people want to spend on a particular chip," he says. Oh and it does change frequency, he says, but only between transmissions. It changes frequency between signals. "You want to keep radio on for as small an amount of time as possible, to save power. The best thing is to robustly find a frequency and send all the data."
Low-energy Bluetooth will use a star topology so one master can only talk to slaves that are in range at a given moment, but there could be very many of these, if low-energy Bluetooth appears in devices Heydon predicts, such as watches, shoes and heart monitors.
Is the star topology a weakness compared to ZigBee's mesh? Not at all, says Heydon. In fact, using a mesh is one of ZigBee's mistakes: "You can't do low power and mesh at the same time," he says, pointing out that any mesh relay node can't go to sleep to save power, as it has to stay ready to hand on data it receives. "The only way for nodes to relay info, is to have them on all the time."
The best way to mesh devices is hierarchy, so masters are linked over appropriate technology, he says: "One technology to do it all is actually the wrong design." Heydon expects a lot of data gathered by low-energy Bluetooth to find its way onto the Internet. Perhaps data gathered by Bluetooth bathroom scales would go, via your phone, to a medical website or - God forbid - to your Facebook page.
Medical systems body the Continua Alliance, is looking at using low-energy Bluetooth for just this sort of application, says Heydon. There's no published list of members of Bluetooth SIG's low-power group, but Heydon reckons it's a big cross-section of all silicon vendors - "not just Bluetooth vendors. There's huge industry awareness of this technology."
There's another advantage Zigbee might claim - it uses a published standard, IEEE 802.15.4 for its radio links. But standards aren't documents, says Heydon, they are what is sold in numbers in the market.
And using that radio doesn't even give ZigBee much benefit. "ZigBee doesn't control the radio design," says Heydon. "The only way to really build a low-power low energy system is to build it form the bottom all the way up to the top. The Bluetooth SIG is the only organisation that is capable of doing that."
Other benefits Heile predicts 600,000 smart electric meters will have ZigBee in them this year, but Heydon thinks Bluetooth has bigger fish to fry: "They can have the electric meters," he says. "You want to control light switches when you are in a chair - and the device you will have with you is your mobile phone."
"The market Zigbee, has been talking about is reducing over time," says Heydon. "The big problem is getting volume and Bluetooth has a big tick box next to it."
It will be a very small additional cost to add low-energy Bluetooth to existing Bluetooth devices, he says, which means the cost of putting it in phones will be close to zero. Silicon companies can start adding it in mid-2009, when the standard is complete, he says, and it should spread rapidly in new mobile phones creating "an instant market" for accessories and home-installed sensors.
This allows other benefits, like presence. For instance, lights can turn on, or your stereo can wake up, when you walk into the room. A car can adapt itself to your preferences when you get in, and a computer can lock itself and maybe turn itself off, when you walk away.
Possibly the most useful thing, though, could be helping find lost items, says Heydon. "If you lose your phone, your watch will tell you where it is."