IT Jobs
Time for a 60GHz gold rush?
If UWB doesn't work, how about another fast wireless technique?
By Peter Judge, Techworld | Techworld
Published: 01:55 GMT, 08 August 05
The radio data world is crowded with grizzled prospectors, all shouting: "There's gold in them thar waves!"
Some are panning for nuggets in the frontier land of ultrawideband, while others are digging frenziedly in the badlands of 802.11n. One lone prospector is gibbering about an undiscovered lode called Xmax, while the rest all shake their heads.
Earlier this month, another 49er mosied into town, from another territory. Sibeam isn't giving away too much about where it's hopes for gold are located, but the rumours are that it's to do with 60GHz signals.
60GHz isn't undiscovered territory - quite the reverse. In 1895, some 30 years after Maxwell's equations outlined the possibility of electromagnetic waves, the Indian physicist J C Bose, who studied under Lord Rayleigh at Cambridge, was channelling 60 GHz signals in his Calcutta laboratory and at the Royal Institution in London. He signalled with them, ringing a bell remotely, and his equipment - much of which still exists - included a wonderfully Victorian polariser constructed from a copy of Bradshaw's railway timetable interleaved with tinfoil.
60GHz today
Today's 60GHz research has the benefit of fast processors. Bose plotted the non-linear reactions of semi-conductors, but he was 60 years too early for any solid state help in applying his research.
The Berkeley Wireless Research Centre, is developing CMOS chips for 60GHz, which could allow low-cost, low-power applications.
The IEEE has set up a task group to create a standard for the physical layer at that frequency. IEEE 802.15.3c is next-door to the 802.15.3a ultra-wideband group in IEEE-space, and is creating a lot of interesting according to its overseer, Bob Heile, who chairs the 802 committee told CommsDesign: "I've been surprised by the number of people attending the meetings. I figured maybe 10 people might attend, but we're consistently getting 20 to 30 at the meetings."
The group is about two years from a standard, but anything it produces could reach the market relatively easily, at least in the US. The spectrum regulator there, the FCC, has allocated a licence-exempt band at from 57 to 64GHz, enough to give up to 2Gbit/s throughput. Any equipment could be sold on the same basis as Bluetooth and Wi-Fi.
The UK regulator Ofcom is consulting about this space as part of its Framework Review, but had no comment to make at this point.
Commercialisation?
With all this activity, is it time to start panning for gold in 60GHz?
The first to stake a claim was NewLans, set up by serial enterpreneur Dev Gupta, discussed at some length in 2004 but now impossible to locate.
Last month, another potential miner broke cover. Sibeam is reluctant to discuss what its actual technology is, but there are several factors linking it to 60GHz.
Sibeam says it is working on Gigabit wireless, and its own press release mentions it has hired Sohrab Emami, one of the BWRC members working on 60GHz CMOs, and James Gilb, technical editor of the 802.15.3c task group.
That's enough for some to conclude that Sibeam is working on short-range 60GHz technology.
... or not?
However, Sibeam's vice president of marketing, John Marshall, is keen to keep the air of mystery: "We've got a fairly broad base of talent that comes from diverse areas of wireless, including 802.11n," he told us. "The real emphasis is we're trying to work on Gigabit/s wireless, and finding the best way to do that."
Other Sibeam people include head of research James Gilbert, who comes from Atheros where he worked on 802.11n, and director of engineering Bernard Shung, who worked on gigabit switches at Broadcom. Marshall is keen to project an image of a wireless powerhouse, ready to push existing Wi-Fi technologies to higher speeds, or bring in new technologies where appropriate.
Limitations
There's a reason for this. 60GHz is interesting, and genuinely can offer Gigabit speeds. However, it is limited by the propagation characteristics of electromagnetic radiation at that frequency.
Like infra-red, 60GHz waves rely on line-of-sight propagation. The infra-red communication standard IRDA, which required two devices to be on a desk looking at each other, was a failure compared to Bluetooth, which allowed one to be in your pocket.
If UWB takes off, we will have a non-line-of-sight technology for Gigabit personal area networking, sold as both ultra-fast Bluetooth and wireless USB. Against this, 60GHz would play the role of "fast IRDA" - very much a supporting role.
If UWB doesn't take off - and this is entirely possible, given the regulatory and standards hassles it faces, and the possibility it will not find a world market - then we will have to get to Gigabits by other means. 60GHz may be one of those, but there are others.
Some 802.11n contenders are already talking about 500 Mbit/s networking on the existing 2.4Ghz and 5GHz spectrum.
If there is a wireless Gigabit goldrush, it's not necessarily in the 60GHz mountains. It looks like Sibeam is hedging its bets, in case it has to stake a claim elsewhere.
.gif)
Add your commentComments