Ultra-Wideband: New Ultra-Connectivity for the Enterprise?

Like predecessors 802.11 and Bluetooth, ultra-wideband (UWB) is poised to become "the next big thing" for wireless connectivity in the enterprise.

This wireless technology, which was first the subject of experimentation more than 100 years ago, boasts gigabit speeds that far surpass Bluetooth and supports significantly more spectrum than 802.11. Plus, it supports quality of service (QoS) necessary for applications like streaming video and audio.

True, formal standards have not yet evolved - two competing versions of the 802.15.3a standard both failed recently to gain enough support to move forward toward IEEE ratification. And even ardent supporters of the technology acknowledge that once such issues are sorted out, the first UWB applications will likely be deployed on consumer applications.

However, those supporters - and many other observers - believe that UWB is so attractive that it inevitably will become a strong asset in the enterprise, not as a replacement for 802.11 but rather as a complement to that technology.

No Technology-Come-LatelyThe origins of UWB date back to Heinrich Hertz's experiments at the end of the 1890s. The technology has been widely used by the military for applications such as radar, surveillance and sensor applications, but it didn't start gathering momentum until early in 2002.

That's when, after a long battle, the Federal Communications Commission (FCC) agreed to make available unlicensed spectrum in the 3.1 GHz to 10.6 GHz frequency band for commercial use by UWB devices.

The FCC ruling included serious constrictions on power consumption, which means UWB's best chance for widespread adoption is the wireless personal-area network (WPAN). That's because UWB offers impressive bandwidth but a relatively short range.

The 802.15.3a Task Group of the Institute of Electrical and Electronics Engineers Inc. (IEEE) has been deliberating on the best technical solution on which to base UWB's short range, 480 Mbps standard. The group whittled 23 proposals down to a pair.

One proposal called for multi-band orthogonal frequency division multiplexing (OFDM). This proposal was backed by Intel, Texas Instruments and by virtually all major consumer electronics manufacturers.This system uses orthogonal frequency division multiplexing (OFDM) to transmit information on each of the sub-bands, where each must be more than 500 MHz, according to FCC requirements. OFDM's strong points include high spectral efficiency, resilience to RF interference and extremely efficient energy capture. OFDM also has a solid track record in other commercial technologies such as ADSL, VDSL and 802.11a/g.

The other proposal, direct sequence code division multiple access (DS-CDMA), was favored by Motorola and by XtremeSpectrum Inc (the former recently purchased the assets of the latter). While generally more widely accepted, this second proposal still was unable in recent meetings to acquire the needed 75 percent of IEEE voters needed to begin on a standards draft.

This spread spectrum, or code-division multiple access (CDMA), technique spreads the transmitted information across the entire spectrum. While spread-spectrum is well established in commercial technologies such as wideband CDMA, it can be complex because it requires RF and analog circuits along with high speed analog-to-digital converters to process the signal. Because no agreement on a draft standard was reached in recent meetings, the fledgling UWB industry now runs the risk of falling behind its original timetable for having standards-based UWB-enabled gear on the market by early 2005.

"What's One More Network?"

Much of the initial interest in UWB has centered on consumer electronics, where the technology can enable applications such as instant downloading of video from a camcorder to a plasma-screen television.

However, the technology will undoubtedly move into the enterprise. For one thing, PC vendors view UWB as the enabler of wireless USB and wireless 1394. This, in turn, will enable them to untether peripherals such as portable hard drives, printers and DVDs."After wireless USB shows up on PCs, UWB will start to make a presence in the enterprise," predicted Roberto Aiello, co-founder, president and CEO of Staccato Communications Inc., a San Diego-based UWB chipmaker.

Most notably, mesh and multi-hop networks would best make up for UWB's obvious shortcomings in range, Aiello said.

"It's really not about UWB, but more about networking," Aiello said. "Corporations have Bluetooth, 802.11a and 802.11b. What's one more network? Everyone understands how to work with wireless links."

A recent report from Gartner Inc. supports this view. The Stamford, Conn.-based research house reported that wireless networking is becoming commonplace in the workplace. By 2007, more than half of large will use at least five different wireless networking technologies, according to the report.

"Enterprises first will get comfortable with UWB in the form of wireless USB," said Stan Bruederle, a research vice president for Gartner. "There, the domain is the desktop, and then perhaps a printer connection [via UWB] will be next."From there, it's a short way to some kind of UWB capability being added to an enterprise-wide network.

"After all, what IT guy wouldn't want to use 480 Mb on the desktop?" asked Jack Sullivan, founder of Emergent Networks Inc., an Austin, Texas-based start-up that is developing embedded UWB solutions for the enterprise. "The real question is 'What is the best way?'"

Enterprise Applications

Enterprise PCs share similarities with home-based PCs and perhaps even some consumer electronics devices, but networking on an enterprise scale is a big leap of faith.

"We're not talking about [UWB enabling] point-to-point [connectivity] in the enterprise," Sullivan explains. "But UWB isn't just wireless USB or 1394, either - it's much more."Instead, Sullivan added, many corporations are looking at IP-centric networks where the introduction of UWB would broaden the architecture and usage. That's particularly true if its embedded into items like phones, printers and laptops.

Even if UWB exceeds enterprise expectations, neither Sullivan nor Aiello predicted it would replace 802.11. Instead, UWB will extend and enhance the 802.11 architecture in the enterprise, empowering applications such as streaming video and real-time collaboration.

"UWB applications are perfect for where density is high but mobility is low," Sullivan said. For example, users typically download files or view streaming video in their offices or meeting rooms. These areas would become what Sullivan described as "pockets of excellence," where UWB's WPAN capabilities offer higher throughput and quality of service.

"This capability isn't needed in, say, hallways," he added.

Dongles may be necessary at the start to ignite UWB, but industry watchers expect adoption followed by embedding to occur so much faster than what 802.11 experienced."The price points were high when Bluetooth and 802.11 used dongles," said Emergent Networks' Sullivan. "Users were not that comfortable with wireless. Now they are. Look at Intel. First it expounded a wireless local-area network (WLAN) solution. Now that view has expanded to envisioning a UWB radio in every Pentium chip."

That day, most likely, will come because many observers expect PC makers to jump on the UWB bandwagon.

"Putting UWB radios in PCs will eliminate the wires from the backplane while adding new functionality and improved capabilities," Sullivan said.

Enterprises Must Develop Plans

On the one hand, enterprises must develop plans and future guidelines for supporting multiple wireless network technologies. On the other, looking ahead isn't that easy, said Gartner's Bruederle, because UWB is some five years from reaching maturity and a lot can happen in that time.According to Staccato Communications' Aiello, the main impediment for UWB in the enterprise is concern about security.

"Nothing else holds back the IT manager," he said.

There was a time when IT personnel installed a new network element and then watched for security problems, but today assuring network security comes first. Aiello said what some see as UWB's biggest negative - its short range - actually could help quell security concerns.

For instance, if a network is configured so access is not granted below, say, 54 Mbps, UWB's range then would be limited to the immediate area. This would significantly reduce the possibility of improper use.

Yet not every enterprise necessarily wants to improve its network, especially if it represents change. While Staccato's Aiello said, "combining an 802.11 network with UWB will create the ultimate capacity for the enterprise," to reach this pinnacle, UWB devices must prove they cannot only work together but interoperate with myriad other wireless devices without interference."Bluetooth and WLANs don't play well together while UWB could eventually be the Bluetooth killer," said Bob Cutler, senior technologist for Agilent Technologies Inc. "UWB will become what Bluetooth should have been."

Peter Meade is editor of UWB Insider, a monthly newsletter that offers perspectives and commentary on the ultra-wideband market. He can be reached at [email protected].