November 06, 2008, 9:47 PM — Ultra-wideband wireless, which underlies wireless USB products, suffered another jolt recently when one UWB pioneer shut down operations.
WiQuest, a fabless semiconductor company in Allen, Texas, closed its doors when it couldn't raise additional funding. Intel's decision to scrub its UWB research, made months ago, surfaced into public view about the same time. And WiQuest rival Alereon announced it had acquired part of Stonestreet One's wireless USB business: its USB drivers.
Last May, Texas Instruments announced it was withdrawing from the WiMedia Alliance, which oversees the UWB specification and interoperability testing, and backing 802.11n Wi-Fi as the next underlying radio technology for the Bluetooth protocols, a project launched in February by the Bluetooth Special Interest Group. That seemed to push the SIG's plan to run Bluetooth over UWB, announced two years ago, to a lower priority.
And earlier this year, Sony announced its own UWB alternative: a short-range, low-power wireless technology, dubbed TransferJet, which has won the backing of more than a dozen of the world's biggest consumer-electronics companies.
On the plus side, start-up USB-chipmaker Staccato Communications announced that its new Ripcord2 family of single-chip CMOS USB silicon devices had won certification from the WiMedia Alliance. The Ripcord2 silicon supports additional frequencies (so vendors can build UWB products using one device that can work globally), hopping among channels (increasing range), sidestepping radar or cellular networks to avoid interference.
UWB targets what are called "personal area networks" or short-range radio connections among client devices, such as a laptop PC and a group of peripherals. Most current products using UWB radios implement the wireless USB standard, intended to replace the snarl of USB cables that interconnect a wide range of clients and peripherals. The range is short: It can reach as far as 60 feet in some implementations, but most users connect devices that are less than 15 feet apart, says Eric Broockman, CEO of Alereon.
UWB's strengths are its very-high throughput -- now typically in the 50M to 100Mbps range with first-generation chips or chipsets -- and very low power, about 1mW per Mbps compared to 15 to 20mW per Mbps for 802.11g and estimated 6 to 8mW per Mbps for 802.11n. Alereon has demonstrated throughput of 200Mbps, according to Broockman, and the next generation of UWB silicon will offer peak rates of 300M to 377Mbps and eventually more. But that's still well below the 450Mbps originally touted by UWB proponents.
The fundamentals of the technology remain solid, even if the U.S. and global economies make it tough for UWB start-ups to find new venture funding, Broockman says. The number of wireless USB products is growing because of the sheer convenience of using wireless connections and the ability to use today's USB-based software transparently.
One stubborn challenge, however, is lowering the cost of UWB chips from about US$10 today to the $2-to-$5 range, in order to trigger mass adoption in the wireless USB and mobile handset markets.
Broockman sees Bluetooth over UWB as a key development. That's because cellular handset makers have big investments in Bluetooth software designed for mobile phones and applications. Now that investment can be leveraged over a much faster UWB link, without the costs and complexities associated with Wi-Fi.
Stonestreet, for example, is continuing aggressively with its work in that and other areas of UWB, according to Broockman. Alereon's acquisition was based on Stonestreet's decision to divest itself its PC driver software, which Alereon was already licensing and wanted to secure, he says. Alereon earlier this year began product development for integrated laptop-based UWB.
Intel launched its UWB design project about five years ago, as part of its new business-initiatives group. According to one report, Intel business units determined they could buy UWB chips cost-effectively from other vendors if needed, rather than creating their own. Intel has investments in Staccato and another UWB start-up, Wisair.
That decision was actually made earlier in 2008, and widely known among OEMs, according to Broockman. "We decided to go after the PC laptop market [ourselves] because we saw they had done all they had intended to do in that space," he says.
According to some reports, WiQuest had been struggling to move from a two-chip to a one-chip UWB product, and didn't yet support the multiple frequency bands needed for a global UWB silicon product.
