Ultra Wideband
When approved by the FCC early last year, Ultra Wideband (UWB) seemed to be the “Next Big Thing.”
UWB could re-use the whole frequency spectrum by sending digital signals under the noise floor (see sidebar). Thus it promised huge hunks of bandwidth for everyone. But there was one small problem: Spectrum owners didn’t trust it. So the Federal Communication Commission’s rules on UWB were strict. There would be no use of UWB at frequencies under 3.1 GHz. And power limits were set so low that the signal could hardly travel anywhere.
Still, a few entrepreneurial companies, such as Time Domain Inc., of Huntsville, Alabama, and Xtreme Spectrum Inc., of Vienna, Virginia, jumped in. Time Domain delivered a chip-set called PulsON, while Xtreme announced a chip-set called Trinity.
The Feb. 14 approval of UWB by the FCC was fairly application-specific. It allowed for the development of imaging systems under Part 15, for collision avoidance radar in the 24 GHz band, and for home networking devices operating between 3.1-10.6 GHz.
The power limits on these pulses, however, were set very low, as much as 10 to 12 dB below the narrowband limits for unintentional emissions. Despite these limits, cell phone, GPS and satellite radio firms objected to UWB’s approval, fearing interference with their licensed frequency. UWB companies also objected, saying the rules were too strict.
Congress held a hearing in June, with House Energy and Commerce Committee Chairman Billy Tauzin, R-La., taking the side of UWB advocates and criticizing the National Telecommunications and Information Administration (NTIA). The Pentagon and NTIA pushed for the limits.
The October release of an FCC study indicating microwave ovens and laptop computers pose more of a threat of interference than UWB devices seems unlikely to mollify the critics.
When the rules were approved, the FCC promised to decide on any reconsideration requests within a year. A spokeswoman told MRT that the deadline expires in February, and “We hope to meet that commitment.”
For now it is up to the industry to live with the limits, to prove the case that the technology won’t interfere with licensed spectrum rights, and to build up the business.
So where will UWB be in 2003? According to Paul Marcik, an analyst with Allied Business Intelligence Inc., Oyster Bay, New York, it will be in Moore’s waiting room, seeking high-cost applications that might start costs heading down.
“How do you get the price down? By getting demand. How do you get demand? With a low price. You need to introduce it first into high-cost items and offer capabilities that aren’t available in any other way,” Marcik said.
So for 2003, large chipmakers such as Intel don’t see the market as holding enough promise to ramp-up production of UWB product.
“We think of this as wireless USB,” said Kevin Kahn, an Intel fellow and director of communications and interconnect technology in Hillsboro, Oregon. “That’s a good way to think about the application space we’re interested in.”
It’s just a way of replacing the wires under your computer with digital pulses.
Because of how UWB operates Kahn says there are three main areas of application:
-
Search and rescue imaging applications. This is actually where UWB started, as a military technology used for things like ground-penetrating radar. The problem is that waivers from the power limits are needed to make this work.
-
Location-based services with low data rates. Because UWB uses high frequencies with short wavelengths, distances can be triangulated precisely when two emitters are used.
Thus, warehouses or large stores could use UWB to take inventory remotely, replacing bar codes with chips and radios.
-
High data-rate, short range applications. Because UWB signals can cover a wide range of frequencies, but the waves themselves are very short, it’s a technology that might replace the wires linking computer devices together in homes and offices. This is where Intel is placing its emphasis.
But such high-volume, low-cost applications are some time away, Kahn warned.
“We still view this as a research project. We’re getting closer to the point where we might put some product initiatives together. But officially we’re still looking at it as a research project.”
High price, high profit
That search-and-rescue market might look small to Intel, but it looks like a fine business to Jeff Ross, vice president of corporate development for Time Domain.
So this month the Huntsville, Alabama-based company is releasing RadarVision, a $29,500 imaging device using UWB for police and fire agencies.
“You slip it on your arm, and it has a screen on the back side. You hold it against or near a wall. It will image for about 30 feet, and update the image every third of a second,” Ross explains.
The screen gives a 120-degree view around the holder.
“It weighs a little more than 10 pounds. It has an antenna array and packaging. We’re doing a lot of signal processing, and it’s designed to be dropped and survive. Most of the weight is in the packaging,” Ross says.
Customers want the weight for the durability.
“There are 14,000 federal, state and local law enforcement agencies in the U.S. and the target market ends up being several thousand agencies,” he said.
The 10-pound device can be used by police to detect criminals and by firemen to detect victims in a ruined building.
RadarVision is also a throwback to the original military mission of UWB. Its original application was in “ground-penetrating radar” and collision avoidance systems. High volumes of radio emissions reflect off a heated object (like a body) and are shown as an orange blob. The Time Domain device uses high-resolution detection to make out shapes.
Much of Time Domain’s work, however, is still tied-up with military contracts, Ross admits, and one contract already in hand will develop an improved device with 3-D positioning.
While RadarVision requires a waiver from the current FCC UWB emission rules, it still puts out just 50 milliwatts of power, 1/10,000th the power of a cellphone.
Big volume needs standards
The key to creating high-volume UWB applications is standards.
Since September, the main forum for this work is the IEEE 802.15.3 committee, with major industry players organized there under the WiMedia Alliance (www.wimedia.org).
The 802.15.3 “Personal Area Net-work” (PAN) standard was originally approved for use in the 2.4 GHz band used by Wi-Fi networks (www.ieee802.org/15/pub/TG3.html). WiMedia is working to get a version of the standard approved for use by UWB devices.
Eric Boll, senior applications engineer with Xtreme Spectrum, says PANs are the big opportunity for UWB.
“The system works by picking up the strongest signal, whether direct or multi-path. That makes it more robust in clustered environments, inside buildings,” he says.
Since the FCC will allow UWB devices to operate across 7 GHz of spectrum, speeds of 100 Mbps within a PAN are quite practical, he adds, within 10 meters.
“You could connect display panels that are remotely situated in the room, simplifying the connections in your home theater and your home video system,” says Boll, even linking consumer electronics devices with home computer networks seamlessly, without wires.
But there is a problem. Because they are putting out very low-power emissions across a broad frequency spectrum, properly-designed UWB antennas are vital to proper operation, says Martin Reynolds, a Research Fellow with Gartner Dataquest in San Jose, who wrote one of the first UWB market reports in February 2002.
“Badly designed products cause trouble,” he says. “Antenna issues are vital in this space.”
If an antenna is broken off and becomes a stub, for instance, UWB energy could be concentrated on one frequency and, thus, interfere with other wavelengths.
“There also may be coupling effects, things located near the antenna that act as re-radiators,” he says.
All these issues must be hashed-out in the standards process, and that will take time, Reynolds says.
Some can’t wait
Not every company is willing to wait until standards are set or the regulatory environment clears up before making big claims and setting big plans for UWB.
At the Western Cable Show in Anaheim, California, in December, for instance, Pulse-Link Inc. of San Diego showed off a UWB-over-wires technology designed to double the capacity of digital cable systems.
President Bruce Watkins says that’s just one part of a two-part strategy.
“Our chip-set solution has two chips – the RF analog is silicon-germanium, and the digital band controller is CMOS. The CMOS is the same whether used for wired and wireless.
“We put equipment at the cable head-end and the customer premise.”
The customer premise equipment (CPE), which can go into a set-top box or an external dongle, “will be a dual cable-wireless chip set.
That wireless chip set, Watkins claims, can transmit-and receive 400 Mbps within a distance of 10 meters, falling from there to 7 Mbps at 100 meters.
To Watkins that makes his system, not a PAN, but a LAN.
“So every home or business that receives this massive new bandwidth our technology enables, will terminate in a 400 Mbps wireless LAN that can touch everything in the home. It’s a Trojan Horse into the home for our wireless. We call it Bridged UWB Architecture (pronounced Bubba).”
What makes this possible, says Watkins, is patented modulating technology and Pulse-Link’s work with antennas.
“It’s not the FCC who said you can only use it for PAN; it’s the way companies out there implement,” he says.
One other advantage for Pulse-Link, Watkins says, is motivated buyers in the cable industry.
Without new technology, the channel capacity of digital cable systems will be cut dramatically when HDTV becomes common.
Pulse-Link not only fulfills that requirement, but puts cable companies into the LAN business.
Chip sets should go out by April, he says.
Revolution revisited
Before the FCC approved UWB a year ago, it was called revolutionary.
After the rules were announced, the revolution was cancelled.
Thanks to a lot of work by a lot of people, however, the revolution is on-again. The power limits may be raised to allow more data throughput.
UWB silicon will roll out of chip foundries this year.
By 2004, the technology’s boosters insist, we could have millions of devices in homes, offices, and warehouses, devices vital not just to extending broadband but to winning the War on Terrorism.
Contacts
Xtreme Spectrum
Eric Boll, senior applications engineer
Diane Orr, public relations
(408) 377-0308
Intel
Kevin Kahn, Intel Fellow and director of communications and interconnect technology, Hillsboro, OR (503) 264-1607 (PR)
Paul Marcik, Analyst
Allied Business Intelligence, Inc.
69 Hamilton Avenue
Oyster Bay, NY 11771
(516) 624-3113 x25
[email protected]
FCC Press Office
David Fiske
(202) 418-0513
Bruce Watkins
Pulse-Link
President
(858) 587-9001
Martin Reynolds
GVP & Research Fellow
Systems and Semiconductors
Gartner Dataquest
251 River Oaks Parkway
San Jose, CA 95134
(510) 657-7067
[email protected]
