A decision at an IEEE meeting this week could bring together two factions battling over a new wireless technology. But it's more likely to drive them to all-out market warfare.

Cut-throat capitalism might prove to be the fastest and most efficient way to create a de facto standard for Ultra Wideband (UWB) wireless networking. Advocates say UWB could become the wireless equivalent of USB to link an array of mobile devices and consumer electronics at distances up to 30 feet and at data rates of up to 480 Mbit/s.

In corporations, UWB could become the main way that notebooks and PDAs connect with peripherals and share multimedia data in an ad hoc manner.

The IEEE 802.15.3a Task Group (TG3a) is charged with crafting a high-speed, physical-layer standard for handling wireless multimedia traffic. The parent 802.15 group is developing standards for so-called personal-area networks, including those based on two other wireless technologies, Bluetooth and Zigbee. Task Group 3a members earlier this year winnowed 23 proposals down to two, both based on UWB. In the last three meetings, about 60 percent have voted in favour of one proposal, from the MultiBand OFDM Alliance (MBOA), while the rest prefer one based on direct sequencing (DS), proposed by Motorola.

Unfortunately, any proposal needs 75 percent to be adopted, so the standard has remained in deadlock since July. "We have two positions that are both claiming to best satisfy the market requirements, and no ability to prove (their claims) one way or the other," says Bob Heile, who chairs the TG3a and 802.15 groups.

UWB's roots go back nearly 40 years. Until recently, UWB has been limited mainly to classified defense communications and to systems such as ground-penetrating radar or wall-penetrating imaging. Then, in early 2002, the FCC ruled that UWB radios could run on a given chunk of public spectrum (3.1 to 10.6 GHz) under strict limits. (read our overview of UWB technology).

Conventional radios, such as those in WLANs, have a single radio signal called a carrier wave that beams over a specified frequency. By contrast, UWB doesn't use a carrier wave: instead it uses short pulses of energy and spreads them over a range of frequencies using well-known modulation techniques such as orthogonal frequency division modulation (OFDM) or direct sequencing. These two techniques are the basis of the rival proposals offered to TG3a. In both cases, advocates say the result is very high bandwidth, very low power, and relatively simple and inexpensive radios.

Competing proposals
MBOA is a group of about 40 vendors including most of the world's biggest makers of semiconductors, consumer electronics and computers.

The MBOA proposal divides the UWB spectrum into at least three bands, and uses OFDM to create numerous, narrow channels within these, and to "hop" between them. "If you break up the spectrum into 500-MHz chunks, it simplifies the (radio) architecture and lets you use CMOS (silicon technology)," says Mark Bowles, vice president of marketing for UWB start-up Staccato Communications, and a co-founder of MBOA. OFDM is touted for its efficiency in capturing radio energy, especially useful when the energy reflects off various surfaces and hits the receiving antenna out of phase, causing interference.

The second proposal, called Direct Sequencing, is based on technology created by Xtreme Spectrum, which Motorola acquired in November. This approach uses a technique called direct sequence spread spectrum, which lets many transmissions share the same frequency ranges.

Among other things, this makes it easier for many small groups (called piconets) of UWB devices to link with each other. Advocates say this approach will cause less interference with existing licensed spectrum users than MBOA's proposal.

Last week at the Consumer Electronics Show in Las Vegas, Samsung demonstrated streaming HDTV signals over a UWB link based on the Motorola Xtreme Spectrum chipset. Samsung also is a member of the MBOA.

This week, members of Task Group 3a are expected to vote again on whether to adopt MBOA. No one seems to think the proposal will gain the needed 75 percent. Task force members' patience seems to be running thin with the lengthy process.

Waste of time?
"If we don't see any headway, I question how much longer companies will continue to send people," says Mark Fidler, senior engineering scientist with the imaging and printing systems group at Hewlett-Packard (HP), and a TG3a participant. HP is an MBOA member. The Singapore meeting in September was a "complete and utter waste of my time," he says.

MBOA plans to publish its own UWB specification next month. Members will start building products based on that specification, says Stephen Wood, who oversees UWB strategy for Intel. "There's very aggressive work being done in MBOA," he says. "And upcoming announcements will reflect this."

Motorola won't sit idle either. "We need to work with the companies that need a (UWB) solution now, before a standard (is final)," says John Barr, director of standards realisation for Motorola. "No MBOA members have a product they can give to customers to run and test."

The IEEE could go ahead with both proposals, but that's not an idea that many like. "Two (physical) layers? That's a cop-out because we couldn't make up our minds," HP's Fidler says.

Heile, a veteran of IEEE battles, seems unruffled. "The folks on the 802.11g (WLAN) group fought like dogs for two years," he says. "We've only been at this for six months. We've got a ways to go."

The fighting won't be an obstacle, he says.

"I hear people whining that this (lack of progress) will retard the market or fracture the market," he says. "The reality is, we don't have a market yet. The US is the only major country that allows the use of UWB without having to register it with the regulatory authority."