Electronic Design

The Year Of Ultra-Wideband

Ultra-Wideband (UWB) is one of the hottest wireless technologies. We’ve been hearing about it since 2002, when the Federal Communications Commission authorized the use of the 3.1- to 10.6-GHz spectrum for unlicensed services. Sooner or later, maybe it will settle down so designers can begin incorporating it into their work.

Lots of chip development has been going on, and some real products are finally available. If it weren’t for the division in the IEEE UWB standards process, perhaps we would have seen some of these products sooner. The IEEE 802.15.3a Task Group split over the technologies and hasn’t been able to work out a compromise yet.

One of these factions, Freescale Semiconductor, bought Xtreme Spectrum’s Trinity chip design and brought it to market. The Freescale design uses direct-sequence (DS) coded pulse-type UWB, which is a more traditional approach.

The other faction comprises all of the other semiconductor companies, which got together and developed the orthogonal frequency-division multiplexing (OFDM) version of UWB. Next, they formed the Multiband OFDM Alliance (MBOA) to promote it. The MBOA eventually merged with the WiMedia Alliance to provide a formal standard that members could subscribe to and to eventually offer testing and certification of UWB radios, a la the Wi-Fi Alliance.

As for the IEEE standards process, forget it. The 802.15.3a Task Group recently voted to formally abandon the effort to establish a single standard. So we’ll have two standards to choose from, the DS-UWB approach and the OFDM version from WiMedia.

While Freescale probably was the first company to market a full-blown UWB chip set, it hasn’t been widely adopted yet. I suppose potential customers were waiting to see what the OFDM WiMedia crowd was going to do. Well, that bunch finally has done it, and quite a few vendors are promoting their new chip sets.

Alereon, Focus Enhancements, Staccato, Texas Instruments, WiQuest, and Wisair all offer the same thing—a chip set and a way to interface with a USB port. WiQuest built its USB port right into the chip, giving the company an advantage. Wireless USB seems to be the sweet spot for UWB. Freescale has a USB interface too. In fact, Freescale may have been the first company to implement USB by UWB.

The OFDM version of UWB uses three 528-MHz wide bands in the 3.1- to 4.8-GHz range. There are 128 channels or carriers per band using quadrature phase-shift keying (QPSK). The max data rate is set to 480 Mbits/s, which is the fastest speed for USB. Higher speeds are possible with additional tricks. WiQuest already has incorporated capability for up to 1 Gbit/s.

The big issue with UWB is range. Top distance is about 10 m, making it a personal-area networking (PAN) technology. Its maximum rate occurs at a much shorter distance, about 2 m. At the maximum range, the data rate will be at least 100 Mbits/s. The standard scales down to 53 Mbits/s. Of course, the rate is a function of the system’s particular environment, including walls, ceilings, floors, and other obstructions. Don’t complain, though, as 100 Mbits/s is pretty fast for almost any application, including video.

Video seems to be the other hot UWB application. There are quite a few forthcoming uses for cable-free video transport from the TV set to a plasma screen, from a DVD player to the TV set, and so on. Eliminating the video and audio cable rat’s nest is an honorable application. But it has to be fast, and UWB seems to be more than up to the task. The range is well within what UWB offers, too. They seem meant for each another.

The video interface may be an IEEE 1394 port, which has been around for a while. Apple calls it FireWire, and Sony calls it i.Link. This super-fast serial technology uses wire or fiber at a rate up to 3.2 Gbits/s. It’s pretty easy to get up to 400 Mbits/s on a CAT5 cable using 1394. Add the UWB, and voila, you have wireless video.

Mike Krell, marketing director for Alereon, said that UWB’s value proposition is its power efficiency. it offers very high speed while consuming extremely little power. It will do its best job with portable devices like camcorders, digital cameras, and video and audio iPods and MP3 players.

As for who wins the UWB standards war, it appears that the WiMedia OFDM group has the critical mass, the momentum, and the most members. With the WiMedia standard recently accepted as the ECMA International UWB standard in Europe, again the OFDM approach is the winner. But don’t count out Freescale and the DS-UWB approach. It will find a place as well. As it turns out, the real competition may not be within the UWB community but with another promising video transport technology, 802.11n.

802.11n technology is the next stop on the Wi-Fi roadmap after 802.11a/b/g. It offers a much faster but compatible version with a data rate to 600 Mbits/s. The 11n technology centers on multiple-input/multiple-output (MIMO) technology that uses two or more (usually three) antennas and transceivers. Using spatial-division multiplexing, multiple data streams are transmitted on the same band, greatly multiplying the data rate.

A rate to 600 Mbits/s is possible at reduced range. A rate of 150 to 180 Mbits/s is more typical of what users will experience at common Wi-Fi ranges—certainly fast enough for video. The great thing about MIMO is that it really takes advantage of multipath reflections rather than suffering from them. It makes transmission more reliable over longer distances. That may be a better advantage than speed.

If the IEEE standards process wraps up soon, the Wi-Fi Alliance can get ready for its testing and certification programs. We then should see some honest-to-goodness 11n products by early 2007. These products certainly will be able to handle video. Therefore, they could be serious competition for UWB. But UWB has a head start, and it already has made some inroads in the video market. The UWB chips also will be smaller and consume less power than any 11n solution, especially since the 11n chips all have two or three radios or more in them.

On the other hand, 11n will be Wi-Fi interoperable and compatible with the many already installed Wi-Fi home networks. It may draw more power, but in a home wired environment, that probably isn’t a big disadvantage. As for speed, UWB probably wins. Yet 11n’s longer range will make it a bit more reliable over a large household. Overall, 11n probably is the better video transport just because of its reliable longer range, which will translate into the quality of service consumers demand.

It is anyone’s guess how this will turn out. Maybe the two will coexist. In a home environment, I wonder how important total interoperatbility really is. It’s an issue if all of your consumer video products need to connect wirelessly, but otherwise it may not be important.

I think the winner will be UWB because of its low cost, low power, and USB and 1394 capability, not to mention again the head start it has. Yet 11n will be a real winner in its own right. It will triumph in the range war and will even find itself embedded in some cell phones, where UWB may never be. You may get your cell phone video over an 11n hotspot.

I can hardly wait to see what the next big wireless technology will be.

TAGS: Freescale
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.