Packet Switching Types: ATM, Frame Relay, TCP/IP, X.25

Transmission: SONET T-Carrier

Services: [3G] [4G] [Bluetooth] [I-Mode] [WAP] [Wireless and packet switching]

Ultra Wideband

Copyright 2003 Archart, All Rights Reserved

http://www.arcchart.com/ (external link)

When it is eventually standardised, Ultra Wideband (UWB) will join the once exclusive club of wireless standards destined to turn the communications industry on it head. However, unlike Wi-Fi, which sailed effortlessly through the IEEE standardisation process, the UWB route to market has not been as well greased. While the IEEE group charged with defining a UWB standard for wireless personal area networking – the 802.15.3a Task Group – has successfully whittled down the 31 proposals submitted at the beginning of the year to just two, both proposals have the backing of an industry giant – Motorola and Intel. The two camps are going head to head in order for their proposal to be selected. If the impasse cannot be overcome at next month’s meeting, the two factions may be prepared to let the market, as opposed to the IEEE, decide the winning standard.

Bluetooth teased the electronics industry with the promise of un-cabling consumer devices, allowing them to connect seamlessly with each other over personal wireless area networks (PWANs). However, the technology never delivered on this vision owing to its slow time to market and relatively low bandwidth. Ultra Wideband will change all this. Capable of connection speeds of 480Mbps over short distances, UWB will act as a formidable cable replacement technology for all kinds of consumer electronics such as camcorders, printers, digital cameras and laptops, and will be able to effectively stream video and audio throughout the home.

According to ABI, UWB electronics and chips will reach 45.1 million units by 2007, with industry revenue of $1.39 billion. Parks Associates reckon that UWB communications systems will exceed 150 million devices by the end of 2008.

The term “Ultra Wideband” is not very descriptive, but it does help to differentiate the technology from more traditional narrowband systems, such as Wi-Fi and GSM, as well as newer wideband systems such as W-CDMA (the European 3G standard). Rather than broadcast continuous signals at a fixed frequency, UWB uses pulse modulation to transmit a high volume of signals in short low power bursts that extend across a wide swath of spectrum. Consequently, UWB systems benefit from very low power radios, consuming less than a milliwatt of power compared to about one watt for Bluetooth and three watts for cell phones. From an engineering perspective, UWB is far less complex to implement than other wireless technologies. In the absence of complex circuitry, semiconductors can be relatively low-priced.

UWB transmissions can comfortably achieve 100Mbps at short distances and proponents of the technology claim that gigabit data speeds are theoretically possible. In the beauty pageant of wireless technologies, UWB is one of the sexiest.

The body charged with implementing UWB in a standard for wireless personal area networking (WPAN) is the IEEE’s 802.15 Task Group. This is developing the 802.15.3a standard that will allow transmission of 110Mbps over ten metres and 480Mbps over one metre.

The group, which started the year with 31 proposals, had whittled them down to one front-runner. This is the scheme being pushed by the Texas Instruments and Intel-led Multiband OFDM Alliance (MBOA) and advocates splitting the UWB radio spectrum, recently authorised in the US by the FCC, into three or seven bands. However, at an IEEE meeting in Singapore last month, while the MBOA proposal garnered the majority of the votes, it failed to capture the 75 percent majority needed for confirmation. The same happened at a July meeting in San Francisco.

This means that a competing proposal, put forward by Motorola, XtremeSpectrum (XSI) and ParthusCeva, is now back on the table. This group advoates using continuous spectrum instead of splitting it into bands.

There are three issues at the root of the 802.15.3a conflict – one technical, another political and the third strategic.

UWB is unique in that it operates in spectrum already allocated to other services. Because of this, it must operate under extremely stringent limits. For example, a UWB transmitter must radiate no more power than other non-transmitting electronic equipment such as a home hi-fi system. However, sending the UWB signal down narrower channels, as proposed by the MBOA, requires extra power and the Motorola camp are suggesting this may breach FCC regulations. Meanwhile, the MBOA point to test results of UWB evaluation systems based on the MBOA proposal at a FCC-registered test facility. These show conformity to regulatory stipulations.

Apprehension over who owns the technical intellectual property required to implement the standard is where the politics comes in. Basing a standard on a single company's technology raises issues of royalties, which the industry is unlikely to tolerate, given the volumes at stake. Neither party are guaranteeing zero-royalty IP, only a "reasonable and fair" licensing policy.

The Motorola/XSI partnership has a strong strategic advantage in that XSI has UWB chipsets available today. Were its recommendation to be approved, it would effectively have the UWB semiconductor market to itself while competitors, such as the MBOA members, played catch-up.

Clearly, this complex web of vested interests means significant barriers will have to be overcome before a common standard can be agreed. On the flip side, Wi-Fi has shown that a standard brought quickly to market with strong support can be very lucrative for the entire industry, and there are signs this will fuel a spirit of compromise when voting resumes. This happens next month when the 802.15.3a Task Group will have to eliminate one of the remaining two proposals, and then try again for a 75 percent majority. If, after two votes the lead proposal still has not achieved 75 percent of the vote, the process resets once more. If neither of the two camps can successfully woo sufficient votes over to their side, theoretically, this could carry on forever. The worst outcome would be if the two factions decide to let the market, instead of the IEEE, decided which becomes the dominant standard. The would not only retard the development of UWB but also the roll-out of WPAN to consumers.

Copyright 2003 Archart, All Rights Reserved

http://www.arcchart.com/ (external link)

Packet Switching Types: ATM, Frame Relay, TCP/IP, X.25

Transmission: SONET T-Carrier

Services: [3G] [4G] [Bluetooth] [I-Mode] [WAP] [Wireless and packet switching]