A power evolution in wireless connectivity for portable devices

Over the past 2 decades we have witnessed the evolution of the mobile handset from a relatively simple product for portable communication, into a multipurpose all singing and dancing tool. Competition between designers is fierce, each striving to incorporate that key differentiator that will pull the end-user towards their product - while also simultaneously obeying the three commandments of portable device design: cost, size and power consumption.

While many things may have changed in how a user interacts with a handset, the issue of power consumption in particular for portable electronic products remains the same. GPS is an excellent example of a technology faced with this challenge. Location based services are arguably a very attractive feature for portable devices, however existing offerings consume far too much power on a handset’s battery life and this is clearly a major turn off for users.

However, wireless connectivity has drastically changed the functionality of a mobile handset. Bluetooth has proven to be the most successful short-range wireless technology in use today, with the volume of handsets employing Bluetooth technology expected to reach up to as much as 840 million devices in 2008. In June 2007 however it was announced by the Bluetooth SIG that a new even lower power wireless technology is to be incorporated into the Bluetooth specification, Bluetooth low energy, previously known by the names Wibree or Ultra Low Power (ULP) Bluetooth.

Bluetooth low energy is set to change the landscape of wireless technologies altogether. This new low power wireless technology can be used to transfer a few bits of data between compact devices and can run for up to ten years on one button-cell battery. That means that ultra-compact devices featuring Bluetooth low energy will have battery life standby times measured in years, not just days or weeks. A whole new class of wireless connectivity is opening up, addressing new market segments such as watches, training shoes, home automation, remote controls and medical sensors to name but a few.

With such an abundance of possible use cases, the market is expected to be enormous. The potential for the application of Bluetooth low energy technology in sports equipment and particularly personal healthcare/medical applications is huge.

Demonstrations of the technology have already taken place this year, for example at Wireless Japan in Tokyo. CSR - the only company currently with silicon ready for the new low power Bluetooth technology - demonstrated Bluetooth low energy inside the company’s BlueCore7 IC in a mobile phone to show how it can interoperate with weighing scales and temperature sensors. The consumer healthcare market offers enormous potential for the new low power technology, and these recent demonstrations have proven to the industry that as soon as the Bluetooth low energy specification is complete, (expected in early 09) the technology will be quickly and easily integrated into handsets.

Similar use case scenarios could include industrial and home automation as well as healthcare. For example, Bluetooth low energy could be used to allow patients in a hospital to have their heart rate, blood pressure and oxygen levels wirelessly monitored allowing them to remain mobile. The consumer healthcare market could significantly improve healthcare for many individuals, with patients benefiting from the continual monitoring and recording of a health issue. This would use internet connections available in mobile phones to connect these devices to web services. Allowing patients and their doctors to monitor their healthcare trends on web sites.

Other applications for the technology could include Fashion area networks that create links from people’s personal items; such as shoes, outerwear, jewellery and especially watches to their mobile phone. For example a user could keep their mobile phone out of site in their bag or pocket and use Bluetooth low energy to send data from the mobile phone directly to their watch to display who is calling and allow the user to accept the call on their Bluetooth headset.

But how does Bluetooth low energy work? What are its core values? How does it compare with standard Bluetooth and where do the similarities end? When standard Bluetooth was first created, it did so under five core values, namely:
· low power
· low cost
· short range
· global standardisation
· robustness

Each of these core values applies to Bluetooth low energy just as much as standard Bluetooth.

Bluetooth low energy uses the same sniff mode with sniff sub-rating that standard Bluetooth uses for its lowest power operation. Where the difference lies is that Bluetooth low energy uses it from the very start of a connection. That is to say every Bluetooth low energy connection is automatically in a sub-rated sniff mode, and therefore is automatically very low power. So while connected, Bluetooth low energy is as optimised as standard Bluetooth when used to its absolute maximum potential.

Addressing the low cost question, Bluetooth low energy can be built using standard off-the-shelf CMOS technologies, with less stringent timing requirements — allowing lower cost crystals (and a lower external bill of materials than standard Bluetooth). In terms of the handset end, the cost of adding Bluetooth low energy technology to a standard Bluetooth IC will be negligible.

Complying with the other core values of Bluetooth, Bluetooth low energy is a truly global technology that can be used anywhere worldwide with no special regulations required for its use and no restrictive rules that need to be obeyed.

Bluetooth low energy is designed to be robust, using adaptive frequency hopping to ensure it can recover from single blocking systems and interference from other frequency hoppers. This is becoming more and more important in the 2.4 GHz band as other technologies also get used for proprietary applications. Bluetooth has proven its adaptive frequency hopping is not only very robust but also friendly to other radios, like Wi-Fi, within the band.

In its original incarnation, Bluetooth was already designed for low power consumption. Bluetooth in its low energy format is optimised for very, very low consumption. How does it do this? For most of the time, Bluetooth devices are not constantly talking with each other and are simply waiting and listening. Most of the time, devices are idle. The longer they are idle the less power they use. Bluetooth low energy uses fewer frequencies than standard Bluetooth to be connectable. So while Bluetooth has a 1% duty cycle when connectable, Bluetooth low energy significantly reduces this to 0.1%.

Also, a Bluetooth low energy device actively tells other devices that it is around. It then quickly listens to see if any other device is interested in talking with it and, if not, will turn itself off for a long time until it wants to advertise its presence again. This means that it can selectively advertise only when needed, for example when it needs to tell another device that a back window of your house has been opened. It also means that the time from detecting something and communicating this fact to another device is incredible small; it is just under 3 ms to connect, transfer the information and then disconnect, thus further lowering the power consumption of the device.

An obvious question is 'why use standard Bluetooth at all?' if the power consumption advantages of Bluetooth low energy are so great. Essentially, it boils down to the needs of the new class of connected devices. Take a temperature sensor for example. The data communication required doesn't require the complexities involved in a normal Bluetooth connection (service discovery, different packet types, etc). A scanning device will just find the device, find out the temperature - and that's it. So the device itself can be simpler, as well as using less power when it does connect.

In comparison, connecting a mobile headset and phone is a much more typical application for standard Bluetooth. Bluetooth low energy does not have enough bandwidth to enable audio streaming and these streaming data applications are not the target market for Bluetooth low energy. Standard Bluetooth will look after the mono and stereo headset connection and large data flows for synchronisation, whereas Bluetooth low energy will allow a digital watch to display the caller’s ID at the convenience of its wearer.

Thanks to the very small incremental impact in terms of silicon area, power requirements, cost and the simplicity of the technology itself, Bluetooth low energy is set to have the fastest uptake of any wireless standard to date whilst benefiting from the mass adoption of Bluetooth, the highly interoperable parent technology already found in billions of consumer devices. It truly is an exciting time right now to be looking at the future of low power wireless communications for portable devices.