Intel outlines its ultrawideband position
This high-speed, low-power method of wireless connectivity has the potential to revolutionise the way that electronic devices talk to each other, but the companies at the forefront of bringing it to market can't agree on a fixed UWB standard.
On one side, backed by Intel, is the Multiband-OFDM Alliance (MBOA) with its Orthogonal Frequency Division Modulation (OFDM) technology. The opposing group is Motorola/XtremeSpectrum. Concern is growing that the two rival and incompatible camps at the centre of the disagreement may pursue separate UWB standards. This could confuse the market and delay acceptance, setting the development of very high-speed personal wireless networks back by years.
ZDNet UK caught up with Intel’s co-director of Wireless Technology Development, Ben Manny, the day after he attended the IEEE 802.15.3a meeting where once again the ultrawideband (UWB) standards process stalled.
What happened at the Albuquerque meeting?
There’s still a deadlock. The way it should work is that you select one proposal, and then you need a majority of 75 percent to move on when you poll the group for concerns about the proposal. During the process of the confirmation vote, if you vote 'no', you need to state what you would accept to change your vote to 'yes'. At the next meeting, the group that wins gets to come back with the changes, and you vote again. If you don’t get the confirmation vote, you go back to before the selection, and we’ve already been through that two times.
Some companies use some of this openness and fairness for purposes for which they weren’t originally intended. You always question where some of the consultants are coming from when they block the process. This is what’s happening.
So what happens next? Do you wait for the next meeting?
There’s so much support for the MBOA. The number of companies backing it is growing, and they’re going off as if the confirmation had occurred. It’s not a formal organisation, just a collaboration of the sort that the IEEE encourages to help with the standardisation. Our current path is to continue developing this, and what makes it successful is people building products around it.
The key to success [of ultrawideband] is will industry adopt it? Can people build products? We don’t want to slow that down. The IEEE spec is just a radio spec, so we’re working with 1394, the Digital Home Group and others, all of whom have work going on with the rest of what’s needed to make UWB work.
There are reports that the MBOA standard can generate more interference, uses more power and is more complex than the alternative. Is that the case?
Our standard is fully compliant with the FCC published regulations -- in fact, it’s below the allowable limit on peak power. We generate less interference than something like a Gameboy or other electronic device. A certain limited set of receivers -- I hate the term, but it’s what we use, they’re called victim receivers -- that are very wideband are more susceptible to the impulse type of noise that our signal can look like. These are things like fixed satellite receivers.
The question is, what is noticeable interference? We do generate more [than the other proposal] in this one case but it’s only when, for example, the transmitter is right in line with the satellite dish. We’ve done simulations, and we wanted to see how well the testing worked, and the simulations matched our early tests. But the tests are inconclusive and ongoing. We’re trying hard to do good science.
[note: ZDNet UK asked again about power consumption, Manny did not address the issue]
The other camp has said that ODFM isn’t so good at meeting differing international requirements…
It’s easy to make OFDM choose not to use bands, if that’s required. Spectral masks [the rules for how much power is acceptable on different frequencies] may be different in different countries, and OFDM makes that easy.
The last thing we want to do is generate radios that cause problems. Most UWB devices will have other radios in them, like 802.11a, and we don’t want to make that difficult. We can change this proposal as we develop it -- but we’re committed to the spirit as well as the letter of the FCC approval.
The advantage of OFDM is much more effective technology -- you get a lot of performance down that route. Does add a little bit of extra cost, but you don’t have the equivalent costs in the receiver. We’ve done analysis, you can really relax the OFDM construction requirements over systems such as 802.11a which also use it.
Will you make the intellectual property in your proposal available for everyone?
It’s clearly our intent to conform to open IP -- the standard will only be successful if people use it. There’s not enough margin on these components to worry about royalties and intellectual-property payments. Intel’s interests are in getting the standard as widely used as possible.
Pat Gelsinger [Intel’s chief technical officer] told us at the last Intel Developer Forum that come what may with the IEEE, Intel intends to continue developing its proposal and turning it into product. Will there be a standards war between you and Motorola/XtremeSpectrum?
The MBOA is going to continue to develop the specification, and we’d like to see the IEEE endorse it but that’s not necessary. We really don’t want to see the standards war. It’s conceivable that it’ll happen.
So you’re against a war, but you’ll continue to develop your standard come what may?
There’s broad industry support, and the broader that support, the more likely it is that this standard will be created.