UWB Standards Shakedown

Some of the companies involved in the standardization of ultrawideband (UWB) technology at the Institute of Electrical and Electronics Engineers Inc. (IEEE) are hoping that many of the 23 (!) specification proposals before the venerable engineering body get dropped before the process goes much further. If not, it will be a very long time before UWB gizmos and widgets will be in the stores, they warn.

However, this is just one of the issues that need to be dealt with before the funky [ed. note: get down!] new radio technology can be brought to market. On top of the 23 proposals, two distinct groups have emerged within the UWB community -- each advocating use of different frequency bands.

UWB works very differently from existing narrowband radios. Instead of using an carrier signal to encode to a bit stream, UWB sends and receives extremely short monocycle (not unicycle) pulses to enable very high bandwidth (upwards of 100 Mbit/s) over short distances, or low bandwidth over long distances.

The aim of the IEEE's 802.15.3a UWB specification is to develop a standard physical layer standard for high-data-rate, short-range, low-power, and low-cost wireless networking technology that will benefit the entire ultrawideband industry and open up new markets for wireless technology.

According to some observers, a proposal shakeout is already underway: A pre-vote task group meeting scheduled to take place in Texas in May will likely see firms with similar interests joining forces and ganging up on the minnows, outcasts, pariahs, freaks, and so on.

They say that most of the 23 proposals for a new UWB standard submitted to the IEEE in March 2003 are set to be eliminated when the 802.15.3a Task Group holds its first "down select" [ed. note: get down select?] vote in July.

“My intuition is that there will be fewer than 15 presentations in May,” wrote Chuck Brabenac, vice-chair of 802.15.3a Task Group, in a note to the group’s email reflector.

Mike Kelly, an 802.15.3a representative from Focus Enhancements Inc.'s (Nasdaq: FCSE) semiconductor division, agrees with Brabenac's prediction. He says the shakeout is inevitable and a positive move. "If there are more than 15 proposals left after the July down-select, the process could drag on. We’d hope for 10 or less proposals to remain to keep on schedule,” he says.

This view is echoed by Steve Turner from Texas Instruments Inc. (NYSE: TXN), who says if there are still 20 proposals on the table after July, the standard is going to take "a really long time." On the other hand, he says "if there are less than 10, that’s not bad."

Turner also warns about the danger of competing factions splitting the Group. "Everyone wants to come out of this with just one standard," he cautions.

Essentially, however, as Unstrung has previously reported, the ultrawideband community already appears split between those firms that are part of the newly formalized UWB Multi-Band Coalition and those that, well, aren’t (see Coalition Cranks Up UWB).

And what makes this even more interesting is that it’s not clear what exactly constitutes "multi-band," nor what its proponents have in common. Texas Instruments, for example, is outside the coalition, yet considers itself to have a multi-band approach.

At issue is what size frequency bands to use and where these bands should be placed within the 3.1GHz to 10.6GHz of spectrum that has been allocated to ultrawideband.

In general, there’s agreement that it’s quicker and simpler to focus on the low-end spectrum between 3.1GHz and 4.8GHz, and firms such as Texas Instruments and Sony Corp. (NYSE: SNE) appear to favor starting out here, and then expanding into the higher frequency bands at a later date.

Others, mainly from the UWB Multi-Band Coalition, which is a loose affiliation of chipmaker Intel Corp. (Nasdaq: INTC) and number of startups in the sector, appear to be in favor of coming to grips with a wider spread of spectrum straightaway, and then optimizing the solution afterwards. They argue that even though this would be slightly more expensive and time-consuming to get off the ground, it would be better in the long run.

And even once all that's resolved, the thorny issues of what frequency modulation and coding schemes to use must be settled before UWB can be commercialized. "The really big question," says Turner, "is how do you utilize those bands?"

Get down!

— Gabriel Brown, Research Analyst, Unstrung