Do we really need another wireless standard? The WiMedia Alliance says yes. Are they right?

Does the world need yet another wireless standard? With consumers still trying to make sense of three flavors of 802.11, with OEMs still grappling with security concerns, and with chip makers mired in major economic slumps, is the market even able to support a new wireless technology?

Apparently, several large vendors believe it can, and have created a new technology alliance to promote a new wireless standard called "WiMedia." While existing wireless standards are primarily concerned with moving standard data types among wireless devices, the WiMedia Alliance sees a niche for multimedia-based, wireless personal area networks, or WPANs.

The WiMedia Alliance was launched in September with the expressed purpose of developing and adopting "standards-based specifications for connecting personal area, wireless multimedia devices." The companies in the Alliance represent a wide variety of interests in various industry sectors, including chips, imaging, and document management, PCs, and wireless networking. The founders are Appairent Technologies, Eastman Kodak Company, HP, Motorola, Philips, Samsung Electronics, Sharp Laboratories, Time Domain, and XtremeSpectrum; other vendors are expected to join the group in the coming months.

The PAN Plan

At its essence, the WiMedia Alliance is hoping to create wireless specifications based on the media access control (MAC) layer of IEEE's 802. 15.3, which is currently still in draft stage. 802.15.3 specifies a Physical Layer (PHY) using 2.4GHz spread spectrum technology, but WiMedia members are hoping to combine the MAC with an Ultra Wide-Band PHY for a super-fast wireless multimedia standard.

With the federal government's recent approval of Ultra Wide-Band (UWB) technology, UWB appears to be the forerunner as the high-throughput wireless technology of choice for connecting devices that need throughput levels that can support video streams. The IEEE's 802.15 Working Group has a sub-group known as Study Group 3a, which is examining the viability of UWB for throughput speeds in the half-gigabit range.

According to Gemma Paulo, senior analyst for industry watcher In-Stat/MDR's Voice and Data Networking Group, Study Group 3a is expected to choose UWB as the Physical Layer (PHY) for the next-generation of wireless multimedia device. "This very high-rate PAN standard is targeted to provide for speeds up to 480Mbps," Paulo said. "Study Group 3a is expected to reach Task Group status in January 2003, which will open the road to the eventual ratification of 802.15.3a."

Though there is no firm consensus yet on what features the final version of the WiMedia standard will include, according to Alliance officials 802.15.3 will probably be the main building block. But there is also some concern that UWB is still too new in the marketplace for any final decisions to be made.

"The lack of standards in the UWB space is one of the issues the industry still needs to work out," noted Paul Reinhardt, the WiMedia Alliance's acting director. "The WiMedia Alliance is building its technical work (profiles, tools, test, and certification programs,) based on top of the IEEE 802.15.3 draft standard which specifies a 2.4GHz PHY The 3a Study Group is working on alternative PHYs (which will most likely be UWB), and as this effort moves from a Study Group to a Task Group, and through the standardization process, the WiMedia Alliance efforts will expand to incorporate this activity."

The Alliance expects the 802.15.3 standard will be available for ratification in the middle of 2003. It hopes products will be available by the 2003 holiday season.

The 802.15.3 draft is appealing to the Alliance for a number of reasons. According to IEEE documentation, there is currently an "unfilled market need" for a means of networking devices within the "personal operating space" where power consumption, cost, and size optimization constraints prohibit the use of currently available standardized devices (like those based on WiFi). 802.15.3 is based on a centralized and connection-oriented ad-hoc networking topology that supports peer-to-peer connectivity and isochronous as well as asynchronous data.

The WiMedia Alliance feels that 802.15.3 will enable multimedia applications that are not served by existing wireless standards like WiFi and the slower Bluetooth. According to the Alliance, the scheduling medium-access technique selected by WiMedia is optimized for wireless multimedia applications; scheduling is generally a better approach when the majority of the network bandwidth is used for peer-to-peer communications, the group says. The WiMedia platform is based on a centralized and connection-oriented ad-hoc networking topology. In other words, one designated device, or "the coordinator," manages the operation of the personal area network by designating when devices are allowed to transmit data.

Given the preponderance of intra-cell traffic common to multimedia home networks, the Alliance says, this results in a net efficiency improvement versus other techniques such as the one used by WiFi (in which all traffic has to go though an access point and eventually gets repeated to its final destination). The WiFi approach works well for wireless infrastructure access (downstream and upstream traffic), the Alliance feels, but is inefficient when supporting peer-to-peer communications.

What sort of devices might be used on a WPAN? The WiMedia Alliance envisions a variety of home multimedia devices, including digital cameras and camcorders, MP3 players, set-top boxes, game consoles, and even HDTVs. "Some of the main devices we envision for WiMedia networking would be the devices in a home entertainment 'cluster' (set-top box, DVR, DVD, HDTV, digital speakers), digital cameras and digital camcorders, PDAs, printers and digital projectors," said the Alliance's Reinhardt.

Outside of the home, Reinhardt added, WiMediabased photo kiosks could easily, quickly, and wirelessly download digital pictures of tourists on the road. "Whether the device is taking advantage of the low power and ad hoc capabilities of WiMedia, or is a 'permanent,' fixed, AC-powered device that takes advantage of the multimedia-oriented topology of WiMedia and offers the consumer easy set-up (no wiring spaghetti-bail), reconfiguration and security, there is great flexibility and advantage available from this technology."

"WiMedia creates a personal area wireless capability not addressed by any current standard," said John Barr, director of standards realization for Motorola and acting president of the WiMedia Alliance. "Our activity targets high data-rate multimedia connectivity with low-power needs, and therefore enables a broad range of applications in the consumer appliance, imaging, and multimedia markets. The brands supporting WiMedia indicate the potential for our work and ultimate acceptance of the solutions."

The current draft of 802.15.3 provides for five selectable data rates (11, 22, 33, 44, and 55Mbps) and three to four non-overlapping channels. The standard is also secure, implementing privacy, data integrity, mutual-entity authentication, and data-origin authentication for consumer applications. There are also several other advantages over other wireless technologies. The standard offers less complexity per node, which means longer battery life, and does not depend on a backbone network (and thus on a PC) for operation.

Assuming that 3a and UWB support are also added to the mix, speeds ramp up considerably. "We believe that 480Mbps is a realistic goal," said Jim Meyer, vice president of commercial products at Time Domain, a UWB developer and WiMedia Alliance member "First-generation products will support 110Mbps at 10 meters and 200-plus at four meters." Meyer adds that such numbers represent effective throughput, not raw data rate as with the 54Mbps of 802.11a, which provides effective throughput of about half the raw data rate. Meyer expects to see the high throughput capabilities in second and third generation chip sets, which will ship "in the 2006 time frame."

Chris Fisher, vice president of marketing for XtremeSpectrum, another USB developer and WiMedia Alliance member, concurs. "The Task Group 3a speeds have not been approved yet, but we're expecting speeds of 100Mbps on the low end and 500Mbps at the high end."

Analysts at InStat believe that UWB offers significant potential as a high-rate PAN technology, and looks to be a shoe-in as the newest wireless technology whiz-kid. According to Gemma Paulo, the September sessions of Study Group 3a were attended by all of the UWB vendors that have announced plans to enter the commercial UWB market targeting the home PAN, including XtremeSpectrum, General Atomics, Time Domain, Pulse-Link, and SkyCross.

The Competition

The WiMedia Alliance views its new WPAN technology as complimentary to, rather than in competition with, Bluetooth and WiFi. But is this really case? In fact, even with an expected speed boost to 10Mbps (still at least a year oft), Bluetooth--while also an ad-hoc PAN standard--does not approach the speeds of UWB and, presumably, 802.15.3. While it uses less power than a device based on WiMedia probably will, Bluetooth is not likely to pose much of a threat: It will remain a low-cost, low-power, and low-performance solution.

But what about interference? "UWB theoretically transmits short pulses across a very wide swath of spectrum, and does this within the noise floor," noted Gemma Paulo. "Theoretically, UWB pulse are so random and low powered, and are transmitted across such a wide swath of spectrum that they are indistinguishable from noise." Again, in theory, Paulo said UWB will not interfere with Bluetooth, since Bluetooth employs a frequency hopping spread spectrum scheme, in the 2.4GHz range. (See the April, 2002 issue of CTR, available at www.wwpi.com, for a brief discussion of UWB technology.)

More problematic is WiFi, which is already exploring QoS and other packet-based prioritization schemes, as well as faster throughput levels. Plus, noted Paulo, WiFi is backed by hundreds of supporters and 802.11b, though it was designed for data networking, has been extremely successful with various file types. "Most industry backers promise that WiFi will indeed be able to guarantee sufficient QoS to provide for robust multimedia support," Paulo said.

As a local and not a "personal" networking topology, WiFi also has a greater range than 802.15.3 which, in practice, will probably be limited to a range of 20-30 meters for the highest throughput rates (probably 30, not the claimed 54, megabits per second), with speeds likely to slow dramatically at the outside end of the 50-meter range. WiFi silicon is also mature (or at least on the way to maturity) while UWB is untested in virtually all commercial and home applications. And, of course, WiFi is simply an extension of the most battle-hardened networking standard of all, Ethernet.

The opinions of WiFi proponents notwithstanding, however, it remains to be seen if QoS enhancements to the 802.11 MAC will enable sufficient throughput and reliability for the wireless home media center that is on the way. Nevertheless, as we have seen, there appears to be room in the wireless space for a number of robust standards, provided that they differentiate themselves effectively by targeting different market segments. And of course that they do not interfere with one another.

In a broader sense, UWB and the WiMedia Alliance represent a challenging new frontier for the technology industry. Together, the two are the initial wave of the first coordinated attempt by high-tech companies to enter the lucrative home media market, a market where no intelligent device (the PC) need live and where control has been ceded, in large degree, to home electronics manufacturers, not computer companies.

For the past 20 years, these manufacturers have created "dumb" yet extraordinarily reliable devices with idiot-proof setup routines and highly reliable performance. It will be a challenge, to say the least, to add the necessary intelligence and reliability to home electronics equipment when each new piece of equipment will, in effect, be a wireless computing device--with all the overhead that such a role implies. The fabrication costs and technical aspects of adding UWB silicon to home electronics devices have yet to be determined. But, at least initially, observers expect the technology to show up in very high-end, very expensive devices like plasma TV screens, where the cost can easily be buried in $3,000-plus price tags. 'The technology is inexpensive relative to the platforms it will go into," said XtremeSpectrum's Chris Fisher. "We're expecting some early demos at CES in January."

Who will be providing the earliest devices? Fisher won't say, but said to expect "a very small quantity of proprietary devices from some big-name customers" by Christmas of next year.

"Consumer research shows that user satisfaction with [products] requiring more than a 15 second wait declines significantly," said Time Domain's Jim Meyer. "Today a 10GB iPod takes about 20 minutes to fill over Firewire with an effective throughput of 100-150Mbps. Next-generation devices like portable personal video recorders will have 60GB hard drives." Meyer noted that, using high-speed UWB, it will take about 17 seconds to download a 60GB drive to a PC. Now that's fast.

www.timedomain.com

www.wimedia.org

www.xtremespectrum.com