Bluetooth's ULP (ultra low power) mode is expected to consume about 10% to 25% of the power of normal Bluetooth operation, which opens to the door to networking with small, battery operated devices and sensors.
Among the prominently mentioned new applications are:
- Health care monitoring equipment such as blood pressure and glucose monitors
- Sports equipment such as watches that connect to sensors on the body to monitor, distance, speed, acceleration and heart rate
- Automotive applications such as tire pressure monitoring and out-of-safety-range alerts (when the vehicle drifts out of its lane, for example)
- Home and office equipment such as remote keyboards and sensors in home entertainment systems
These applications require ultra-low-power operation because they will run off batteries that may not be replaced for months or years, or, may be integrated into systems that will be discarded when the battery runs down.
In each case, the remote ULP device will communicate with a host device using what are primarily Bluetooth profiles and protocols, although there are ULP extensions. In most cases the host device could be a PC, cell phone, or, watch. But it may also be some new device designed for communicating with Bluetooth ULP devices. Custom host devices are likely to be created for medical applications .
Although it was not prominently noted in most news reports, the networking capability is intended to move beyond PANs (personal area networks). Once data from the ULP sensor reaches a cell phone or PC, it can be channeled into the IP network using GPRS, 3G, WLAN or other protocols that reside on the host device.
Bluetooth ULB devices are expected to run for many months or even years on standard coin-cell batteries. (Power consumption is minimized by combining low duty cycle operation with an efficient radio and streamlined protocol).
Typical stand-alone Bluetooth ULP operationssuch as a watch communicating with a heart rate monitorrequire duty cycles of less that 1%. When transmitting, the chip's power consumption will be less than 15 mA. It will drop to around 30 microA in standby mode, and to 900 nA in sleep mode. A Bluetooth ULP chip is likely to be housed in a 44-mm QFN package.
Data bandwidth will be about 1 Mbit/s and range will be up to 10 meters. The 1 Mbit/s data rate offers an excellent compromise between transmit power (which goes up with
bandwidth size) and duty cycle (which goes down) for a given amount of data.
The radio link has much in common with Bluetooth but the protocol is simpler, and the power demand much smaller.
For background on Bluetooth and Wibree check out these articles:
Fundamentals of the Bluetooth protocol
WiBree Quiz: Five easy facts
In addition to defining a ULP mode, the Bluetooth SIG will also port the Wibree dual-mode specification into its body of documentation. In terms of silicon, this means three kinds of chips: standard Bluetooth (which could be Bluetooth EDR) devices; a stand-alone ULP devices; and, dual-mode devices.
A dual-mode chipset will reuse the Bluetooth RF section with no changes; there will be some additional functions in the Bluetooth baseband section, and a separate protocol stack. The RF link will be at 2.4 GHz, just as it is for Bluetooth. It will also have the Bluetooth RF power level of 0 dBm, and a similar range of around 10m. Modulation will be GMSK (Gaussian minimum shift keyinga small change from the Bluetooth case).
Dual-mode devices will undoubtedly attract a lot of interest because they will be capable of operating in either standard Bluetooth mode or ULP mode. Interestingly, however, the power consumption of a dual mode chip is expected to be about 80% of that of standard Bluetooth.
The dual-mode chip's power performance is lacking mostly because of Bluetooth's complicated but robust modulation scheme, which requires it to hop around in the spectrum. This makes it difficult for the device to go to sleep because it could not follow the hopping. It needs to remain in a semi-aware state at all times.
Hence, higher power consumption. But the Wibree Forum (now subsumed by the Bluetooth SIG) estimated that the dual-mode chip will cost only about 10 cents more than a standard Bluetooth chip.