Discount computer cables

Byline: Dan O'Shea

Ultrawideband technology, the so-called personal area networking protocol, has suffered through a still-unresolved standards dispute, concerns about chipset costs and the perception that competitive technologies are fast encroaching on its turf. And that's all before the first UWB products have come to market.

These challenges have made it easy for some people to dismiss UWB as the next Bluetooth, a previous PAN technology that has struggled to find broad market adoption - or worse, as a technology that will be ineffectual, with semiconductors available but no one willing to buy because there won't be one clear standard.

Despite the negative advance publicity, the real test for UWB is only just beginning and will play out to a larger extent in 2005. UWB developers such as Freescale and Wisair recently launched their chipsets, while other developers, including Pulse~Link and Alereon expect to be producing chips in the coming months.

Those chips are intended for a variety of consumer electronic devices - everything from TVs, set-top boxes and PCs to MP3 players and cell phones. UWB technology will enable these devices with short-range broadband communication, to the tune of at least 100 Mb/s, but eventually exceeding 400 Mb/s - and possibly as fast as 4 Gb/s, depending on the technology, range and application. The potential applications include video streaming from set-top box to TV; downloading photos or video from cameras; providing wireless USB connections to printers and other peripherals from PCs or cameras; downloading MP3s and other files to digital media players; or creating high-bandwidth connectivity between virtually any enabled consumer electronic devices.

Ultimately, the description of UWB as a PAN technology may be somewhat misleading because some of those applications involve inherently mobile devices, and others don't. The solution's most fertile breeding ground lies in the increasingly dense landscape of video, audio and computing devices otherwise known as the modern household.

"It's really a cable replacement technology," said Eric Broockman, CEO of Alereon. "It's for the applications in a room where you would use a USB cable or a 1394 cable. It makes for good cocktail party conversation. Everyone can relate to the problem of having too many cables."

Mutual understanding is something consumers might share in regard to UWB, but at the vendor level, the technology has proved a more divisive force. For more than two years, chipset developers have been at odds over two different technologies vying to become the IEEE 802.15.3 standard.

Multiband orthogonal frequency division multiplexing (OFDM) follows in the footsteps of other OFDM modulation methods, such as 802.11 Wi-Fi and 802.16 WiMAX. Ideally, it would use spectrum in the 3.1 GHz to 4.8 GHz range, which falls between frequencies currently used for in-home connectivity of devices such as Wi-Fi modems and cordless phones.

Within its spectrum, multiband OFDM transmits data over separate carrier channels, each roughly 528 MHz wide. Transmission along these separate paths allows for greater spectral efficiency and low device power consumption, while attaining rates up to 480 Mb/s, according to a white paper produced by the MultiBand OFDM Alliance, a group of more than 130 companies that are backing multiband OFDM as the standard for UWB.

The other potential UWB standard is direct sequence technology, which uses a single, broad carrier channel over which to generate pulses at the speed of about 1 billion pulse per second. Direct sequence is touted by the UWB Forum, which has just under 100 members, including Motorola semiconductor spinoff Freescale Semiconductor.

Martin Rofheart, director of UWB operations at Freescale, said the realistic data rate for both technologies currently lies in the 110 Mb/s to 220 Mb/s range, but added that the road map for direct sequence is to attain rates up to 4 Gb/s, by targeting very short-range applications of less than 10 meters.

The 802.15.3 standard debate has become quite heated at times, with some companies saying the ongoing debate and other factors have led to delays in the standards process. At an IEEE meeting in mid-November, direct sequence won support from more than 50% of attendees during two key votes - but was still well short of the 75% needed to make it a clear choice for the standard.

Freescale's Rofheart doesn't think there needs to be a single standard.

"We think you can have two phys [physical layer protocols] that aren't coordinated together," he said.

Alan Varghese, principal analyst of semiconductor research at ABI Research, thinks that is a mistake.

"This implies UWB devices will not interoperate, so the consumer better know the difference between direct sequence and multiband OFDM. This is not good news."

However, Freescale has proposed that proponents of the two technologies compromise by agreeing that they can have two unconnected phy specs as part of the IEEE standard. Freescale and other companies further propose that the standard include a common signaling mode to allow devices based on both direct sequence and multiband OFDM, as well as any technologies developed in the future, to co-exist within the same network coverage.

Rofheart said he expects some kind of standards agreement by the second quarter next year.

Companies on the multiband OFDM side also don't seem too concerned about standard spats, partly because they think multiband OFDM will dominate.

"It's always preferable to have a single standard, but direct sequence is giving Freescale a time-to-market advantage that is about a six-month window," said Alereon's Broockman. "I don't think a lot is going to happen [in that six months]. Some companies in the Multi-Band OFDM Alliance will ship direct sequence products out of a 'gee whiz' attitude, but I think Freescale will eventually do multiband OFDM. There have been proprietary modem standards in the past that always get run over by industry consensus."

Still, Varghese said standards also lead to cheaper technology costs, and he pointed out that despite earlier hopes, UWB chipsets would be available for as little as $3, Freescale's chipset goes for $19.95. He also said chips coming out next year from MBOA member companies probably will be offered at a discount from that level, but won't be cheap enough to encourage broad adoption.

However, Serdar Yurkahdul, director of marketing communications at Wisair, described the price problem as "transitory." He said chip prices will quickly decrease as supplier choices and volumes increase.

"This is the kind of thing that happens with new technology," he said.

While UWB developers don't think the future of their innovations will be hurt by the standards dispute, UWB also appears to be just one of several technologies vying for attention in a crowded market for home network connectivity solutions.

One of the contenders is something of an incumbent. Wi-Fi already is prevalent in many homes. Though the current maximum bandwidth of Wi-Fi technology is 54 Mb/s, emerging standard 802.11n promises to boost bandwidth to more than 100 Mb/s, with an evolutionary path up to 500 Mb/s. But UWB developers say the applications for the two technologies will differ in some respect.

"I'm not concerned about 802.11n," said Alereon's Broockman. "It's just a natural extension of Wi-Fi, which was just another evolution of Ethernet. They will be complementary. One way to think about it is that Wi-Fi will be crossing rooms, but UWB will be in the room." While 802.11n might be used to network PCs and laptops in multiple rooms of a home to a single broadband Internet connection, UWB will be handling shorter connectivity needs more centered around transferring content from one source to another.

Other technologies with their sights set on home network applications include Bluetooth, the PAN technology that was originally developed for applications within the context of wide area networks, and Zigbee, which, like UWB, is still in the stage of standards development and early market release. But, Bluetooth has struggled to leverage its first-to-market advantage. Meanwhile, Zigbee is more of a sensor technology that is being aimed at applications such as in-home environmental system control, according to the Zigbee Alliance.

Even though UWB faces hurdles on its way to becoming a dominant home connectivity solution, the thing that ultimately will give it the best chance to succeed is early acceptance by dozens of major manufacturers of consumer electronics devices. Alereon's Broockman said UWB-enabled devices will begin appearing on the market in late 2005.