The ultrawideband trade group, the WiMedia Alliance, is shutting its doors, but the group's chief claims that's due to its basic success. A technology transfer will let USB and Bluetooth groups work to bring UWB for high-speed, short-range networking fully to market.
The WiMedia Alliance, a trade group for ultrawideband (UWB) development, announced today that it will transfer ownership of its specifications for physical and media layer technology to two partner groups, and then cease to exist.
The shuttering of the WiMedia Alliance will result in the two trade groups that control USB and Bluetooth taking ownership of the development process: the USB Implementors Forum and the Bluetooth SIG; the Wireless USB Promoters Group within the USB-IF will also be involved.
At first glance, this looks like yet another blow to UWB, which has suffered the shuttering of several firms in the last year focused on the technology. Mostly recently, Tzero, a firm focused on consumer electronics video streaming, halted its principal operations.
However, that might take the short view. UWB's development has continued even as it has failed to take hold in the market with expensive early products. The cost of inserting UWB technology into a device has finally dropped to a level once predicted for years earlier, and which might mark a tipping point—if a tipping point ever occurs.
WiMedia was formed to ensure a common radio platform could serve multiple standards, meaning that one radio could allow TCP/IP networking, Wireless USB, Bluetooth, and perhaps other overlays. The physical layer work has been completed for some time, with various levels of certification underway for two years. The media layer work for standards that talk to the radio, like Blueooth and Wireless USB, can continue within their respective trade groups.
Stephen Wood, WiMedia's president and an Intel technology strategist, said that "it's essentially the same folks in all three groups with very little difference," and WiMedia members' engineers were "spending ungodly amounts of time" in meetings, as well as paying fees for all the organizations.
The decision for the group to disband, Wood said, came from its analysis that there won't be a future divergence in the radio standard, because none of the parties in the various groups want to reinvent the wheel. Such an effort would be extremely costly, and equipment makers (many of which belong to one or more group) have little interest in multiple radios achieving essentially the same purpose.
Wood noted that with the USB and Bluetooth groups, "the members themselves are the coordinating mechanism; they're taking information and cross pollinating it." It's rare that any organization puts itself out of business, but Wood makes a compelling case for his group's demise.
Mike Foley, executive director of the Bluetooth SIG, said, the transfer should "make it easier for members to incorporate into a Bluetooth product." He noted that a variety of intellectual property (IP) issues would need to be resolved, because the SIG requires any specification includes a royalty-free IP license for its members.
How we got here
UWB's appeal was that it was a super-fast (but super-short-range) technology for personal area networks (PANs), including file transfer and synchronization, marketed as a low-power and lower-interference alternative to WiFi. UWB in the WiMedia form was tied closely to Wireless USB as a cable replacement, and to Bluetooth as a high-speed future expansion of that standard.
The technology faced many hurdles. Early battles between competing approaches in the IEEE 802.15.3a PAN networking task group resulted in that group disbanding after years of squabbles. The leading faction, the Multi-band OFDM Alliance, eventually became the only UWB approach remaining, and merged its radio work (on the PHY layer) with the WiMedia Alliance, which focused on the media (MAC) layer.
When UWB devices first starting coming to market, they were incomplete or underfeatured. A couple of early, precertification wireless hubs had an awkward and expensive combination of a dongle, Windows-only driver software, and a multiport USB hub that required AC power. It was hard to see the advantage over wired solutions.
Over the last 18 months, a trickle of Wireless USB gear using UWB has hit the market, including certain Dell, Lenovo, and Toshiba laptops—and cheaper hubs—but nothing has materialized in peripherals or gadgets, even for the early-adopter market.
Wood said that despite this, prices have continued to fall for UWB chipsets, approaching what he terms the "magic price point," which he estimates at about $5 or less to a manufacturer. Historically, he said, that price marks a flashover point at which new technology is adopted. Bluetooth dropped to a $5.00 to $5.50 range before it went mainstream, he noted; UWB chips are "somewhere between $6 and $7" but a large drop from the $15 at introduction.
"We're not surprised we've not seen widescale adoption" because of that price, Wood said.
Wood, involved with UWB technology since its early days in IEEE 802.15.3a, said of the delays in the technology getting to market, and the dissolution of the alliance, "I would have been thrilled if we'd gotten it done faster."
However, he said, turning UWB from a specification into a shipping, legal product involved an enormous number of new challenges: getting spectrum authorized worldwide, a relatively recently completed development; developing a new radio architecture; and producing a new media access control layer on top of that. "When we look back at everything, it has taken us this long to develop this sophisticated a radio," he said.
Wood remains surprisingly bullish for a president about to lose his constituency, but he maintains that the future of the "three screens" convergence—mobile devices like phones, televisions, and computers--necessitates a robust, high-speed PAN standard as part of the mix. "The market's evolving: we were in a state where the PC and the settop box and the phone were extremely different market segments. We're moving into a state where they're the same segment," he said.
"The mobile device is the common element that communicates between them," he added, referring to this ecosystem of devices. "They're all going to end up dealing with the same platform, the same applications, the same general requirements."
While streaming video doesn't fit specifically in either Bluetooth or Wireless USB, Wood said that this convergence would require both groups to have such standards. (The only firm actively pursuing its own video standard over UWB was Tzero. See "Cutting the cord: the state of wireless HD video links," for the rest of what's happening in wireless high definition.)
Ultimately, Wood said that both UWB and in-process 60GHz networking would likely be paired for their complementary characteristics: UWB for interference resistance and penetration; 60GHz for its line-of-sight high throughput within a single room or space.
The Bluetooth SIG, meanwhile, has developed a parallel effort for faster rates using 802.11 technology for bursting file transfers. The Bluetooth 3.0 + High Speed standard, expected to be approved by the SIG by the end of April, works with two compatible Bluetooth modules. The two coordinate a switch to 802.11 (not technically Wi-Fi, a certified version of 802.11 standards) to move large files.
Once a file-transfer is complete, the devices coordinate a switch back to the slower and backwards-compatible 3.0Mbps rate of earlier Bluetooth releases, which consumes less power and bandwidth. Chips with Bluetooth 3.0 + High Speed will be out later this year, with devices incorporating the technology expected in volume in 2010, the Bluetooth SIG's director Foley said.
Source: ars technica
