The Bluetooth Special Interest Group has already, in effect, saved the life of its own technology by agreeing to base future releases on an UltraWideBand (UWB) physical layer. This will not only boost the performance of Bluetooth and so the range of applications for which it can be used, but also avoid an almost certainly losing head-on battle with key UWB technologies such as WiMedia. The SIG has now extended this logic and will work with a range of short range wireless platforms, including Wi-Fi. Such thinking is very positive, pointing the way to a truly unified personal area network (PAN) in which common physical layers drive massive chip volumes and therefore very low costs, while the specific strengths and benefits of individual technologies like Bluetooth are preserved in the upper layers, avoiding the ‘lowest common denominator’ factor that is often the corollary of standardization.
The WiMedia Alliance, whose platform was recently approved as a standard by ECMA and which looks set to dominate the UWB market, has been talking in terms of unified PANs for some time. Its protocol adaptation layer sits between the WiMedia physical/MAC layers and the applications layer, in order to support a variety of technologies ‘ including Wireless USB, 1394 Firewireless, and in future Bluetooth and pure IP. ZigBee, the 802.15.4 standard for low data rate PANs, is also considering a future release based on a UWB physical layer. This work, combined with the Bluetooth SIG’s collaborations with Wi-Fi and the contactless chip technology NFC, is making a unified PAN with multiple applications a real possibility, an outcome that could have a dramatic effect on accelerating the move to ubiquitous wireless. Vendors such as Intel have grand visions of the day when every device and even every human will communicate and access data using tiny sensors and short range, ad hoc connections, but these visions will only become reality with ultralow cost equipment, which means huge volumes to drive chip prices below $1.
As well as pushing down costs, greater integration between different PAN technologies will boost the quality of applications in a sector that has to span a vast number of different uses from factory automation sensor networks to fast inhome digital media systems and, in future, such developments as ‘body area networks’ creating ad hoc links between humans.
“Identifying synergies between wireless technologies and leveraging the experience and knowhow of mature technologies will lead to better wireless solutions, faster,” said Michael Foley, executive director of the Bluetooth SIG.
For instance, NFC (Near Field Communication), a Philips/Sony-backed standard that operates over very short distances of a few feet, has mainly been used for payment systems, where the user waves a smartcard or phone at a terminal in order to purchase items such as transport tickets. But it could also be valuable for instant device pairing ‘ allowing two Bluetooth users, for instance, to identify one another and make a secure connection in milliseconds though an NFC link.
The main barrier to this harmonized picture will not be a shortage of applications but the politics of getting different bodies to work together, and shoehorning the processes of organizations such as the IEEE into a radically new approach to standards and networks. However, to counter this, there will be significant vendor support, especially from those that play in several different PAN markets. Companies like Intel and Broadcom want to drive PAN usage because it, in turn, drives uptake of more and more devices that incorporate their high end chips. They are not interested in the commodity PAN chip market per se, and will be happy to leave this to the volume manufacturers in Taiwan and elsewhere. Companies that may get squeezed in the middle are those that have made their fortunes out of one PAN technology but will not be able to match the price points of those players, and so manufacturers like Bluetooth leader CSR or ZigBee specialist Ember could find themselves challenged.
Other technologies will start to be incorporated into the PAN super-platform over time, adding further functionality and breaking down the old system where PHY, MAC and application layers were specific to each standard rather than being mixed and matched. One to watch is technology for the 60GHz ‘millimeter wave’ spectrum, which can support gigabit speeds for high speed multimedia applications over short distances. Companies such as IBM and Intel are working on reducing the cost of equipment for these high frequencies by creating CMOS chips, while the IEEE now has a standards taskgroup, 802.15.3c, focused on a 60GHz physical layer for its 802.15.3 PAN standard. Similarly, the WiMedia technology could be applied to a 60GHz band for ultra-high speed performance once cost and power issues are addressed.
One of the start-ups focused on 60GHz PANs, SiBeam, has recently joined the EWC (Enhanced Wireless Consortium), the group aiming to set the standard for high speed Wi-Fi (see separate item). This not only indicates the increasing interest in bringing a formerly blue sky technology into the Wi-Fi, UWB and WiMAX mainstream, but the accelerating trend to try to combine the best of all the short range platforms in a single effort ‘ not only leading to more usable and economic networks, but saving on duplicated R&D costs too.
SiBeam, which emerged from stealth mode with a $15m Series A round of funding in June, is part of the 802.15.3c taskgroup and, like Intel and a France Telecom project, is working on making high frequency chips in standard CMOS (most 60GHz chips are currently made in expensive gallium arsenide).
And the mobile operators, one of the key customer bases at which the PAN developments will be targeted ‘ along with the PC makers and consumer electronics giants ‘ are starting to take keen interest in combined technologies too. SK Telecom recently launched a Samsung handset in Korea that acts as a hub for many other devices around the home, such as MP3 players or video recorders, linking them all using Bluetooth. In future, SKT wants a range of different PAN technologies to be available so that it can adapt its hub concept ‘ also being aggressively pursued by Nokia ‘ to different scenarios.
Of course, Bluetooth and Wi-Fi have been coexisting in high end devices for some time, and companies like Pulse~Link and Intel are contributing to the drive for cognitive radios, that adapt to the best available link, regardless of its protocol. This particular trend, though it will be slow and complex, will eventually make the unified PAN a fait accompli anyway, though it will be a long time before such radios can achieve the low price points promised by the more pragmatic mix-and-match approach of the Bluetooth SIG and WiMedia Alliance.
BOX: Wisair
One of the flock of UWB chip start-ups, Wisair, has released its Wireless USB reference design. The Israeli company says the WUSB Hub operates at up to 480Mbps and is based on its second generation UWB chipset with a small form-factor integrated antenna. It coexists with 802.11 a/b/g and Bluetooth devices.
The company is first targeting ODMs but plans to go after the cellphone manufacturers as well.