The 1980s and 1990s saw UWB companies receive generous funding that resulted in successful applications and persistent governmental lobbying, but neither brought about commercial acceptance. Most notably, UWB lacked three crucial boosters:

• spectrum allocation by the FCC;
• creation of an industry standard;
• simple, easy implementation.

The first issue was resolved on Feb. 14, 2002, when the FCC allocated 7,500 MHz of unlicensed spectrum for UWB devices for communication applications in the 3.1 GHz to 10.6 GHz frequency band. The move, which legitimized the UWB market, also fueled the urgency behind formation of an industry standard.

While the UWB spectrum can be used by single band impulse radios built over previous generations, soon a movement began to create an updated approach that would maximize the efficiency and technical prowess needed for today's and tomorrow's high performance commercial applications.

Many turned from single band to a multiband solution that increased performance by encoding information si multaneously in multiple RF sub bands. That approach has emerged as the IEEE 802.15.3 standard's best option for low cost, low power consumption with very high data rate (480 Mbit/second) wireless connectivity among multiple devices in the wireless personal area network (WPAN) space.

As has been learned from previous wireless technologies, such as cellular, 802.11 and Bluetooth, adoption of a standard is a must for reaching omnipresent adoption. Standards organizations (such as the IEEE) and industry groups (such as the USB Implementers Forum and the WiMedia Alliance) are dedicated to ensuring that UWB meets FCC guidelines, which focus on delivering performance that goes far beyond 802.11 and Bluetooth.

Multiband origins

To fuel multiband momentum further, in October 2002 four companies General Atomics, Intel Corp., Staccato Communications Inc. (then called Discrete Time Communications) and Time Domain Corp. created the Multiband Coalition. Philips and Wisair joined the club at the first public appearance in January 2003. Banded together by their common belief that the UWB standard should be based on multiband technology, the coalition worked to refine its united vision, create a merged proposal and repeatedly test its performance. In January, the Multiband Coalition presented a demonstration to major consumer electronics makers, including Eastman Kodak Co., Panasonic, Samsung Electronics Co. Ltd. and Sony Corp., along with big time semiconductor makers, including Infineon Technologies AG, Motorola Inc., STMicroelectronics NV and Texas Instruments Inc. The demonstration gave the group additional backers and serious traction.

The six coalition members saw that their audience shared a common goal: securing an industry standard that would help produce the best possible physical layer (PHY) specification. Out of that gathering came further cooperation, performance testing and soul searching among industry participants, with additional companies joining the coalition especially after the original vision took an unexpected turn. In March 2003, TI presented a proposal, based on orthogonal frequency division multiplexing (OFDM), whose available range was superior to all other proposals at the time.

Some members of the coalition started to collaborate with TI to validate the results. Staccato duplicated TI's system simulation and independently analyzed the system's complexity, presenting the results at the IEEE meeting in May. Further discussions followed, involving all coalition members. Finally, at a meeting in Demver and a follow up in San Diego in June, the decision was made to merge all the ideas and to adopt OFDM as the preferred modulation scheme. The decision confirmed the coalition's resolve to seek the best technical solution for the market.

In June 2003, the Multiband Coalition was reborn as the Multiband OFDM Alliance (MBOA), with Hewlett Packard Co., Microsoft Corp., NEC Electronics Corp., Panasonic, Samsung, STMicroelectronics, SVC Wireless, TDK, Texas Instruments and Wisme joining hold overs, including Femto Devices Inc., Focus Enhancements Inc., Fujitsu Ltd., General Atomics, Infineon, the Institute for Infocomm Research, Intel Corp., Mitsubishi Electric Corp., Royal Philips Electronics NV, Staccato, Taiyo Yuden Co. Ltd., Time Domain and Wisair. The final proposal submitted at the IEEE in July 2003 was also co authored by Sony.

While their backgrounds may be varied, hailing from consumer electronics, digital imaging, home entertainment, personal computing and semiconductors, MBOA members share common interests in designing and building systems as well as building silicon based on UWB, in both single band and multiband modes.

The MBOA believes a multiband OFDM based approach is the best technical solution for the IEEE's 802.15.3a physical layer (PHY) specification, since it satisfies the broadest portfolio of applications and consumer desires.

Introducing Multiband OFDM

The basic concept behind Multiband OFDM divides spectrum into several 528 MHz bands, with each occupying more than 500 MHz at all times to comply with FCC regulations. Dividing the spectrum into smaller bands is independent of the modulation scheme adopted.

The band plan includes four groups;

• those intended for first generation devices (3.1 GHz to 4.9 GHz);
• those reserved for future use at 4.9 GHz to 6.0 GHz;
• those intended for devices with improved SOP performance (6.0 GHz to 8.1 GHz);
• those reserved for future use at 8.1 GHz to 10.6 GHz.

Information is transmitted via the 128 subcarriers in each band, using QPSK modulation. With a contiguous set of orthogonal carriers, the transmit spectrum always occupies a bandwidth greater than 500 MHz.

This method best utilizes the frequency diversity while delivering needed robustness against interference and multipath even in the most challenging channel environments. The signal is sequenced across three or seven bands, with 9.5 ns switching time.

A multiband OFDM system's radio architecture is similar to other conventional OFDM systems. This is a plus for OFDM since it enjoys an outstanding record with other standards organizations, having been adopted by ADSL, 802.11.a/g, 802.16a, digital audio broadcast and digital terrestrial television broadcast. The time frequency code is unique for each piconet as it is changes from one superframe to another to randomize multipiconet interference. The result is a data rate that ranges from 55 Mbits/s to 480 Mbits/s.

The data rate is scalable over several parameters:

In frequency scaling, either three or seven band systems are defined. The three band system is a must for interoperability between different mode devices.

For complexity scaling, necessary data rates (110 and 200 Mbits/s) only need a single D/A converter and mixer for the TX chain for reduced complexity. The digital section scales with future CMOS process improvements, since implementers can trade complexity for performance.

Power scaling requires a half rate pulse repetition frequency approach, which can increase off time to enable power conservation. Here, implementers also can trade power consumption for range and information data rate.

Voting on a standard

On July 25, the majority of voters in the IEEE's 802.15.3a Working Group backed the MBOA's proposal for the standard. Despite that victory, the IEEE requires approval by 75 percent of its voters for confirmation, so the MBOA has additional consensus to gather. Nevertheless, to have re duced the number of proposals from the initial 23 to one in less than six months is undeniable evidence of the swiftness with which UWB is moving as global consensus emerges around a single solution. Once the standard is in place, UWB component makers must be ready to produce large volumes of chips for the big name consumer electronics makers. According to market research firm Allied Business Intelligence Inc., UWB electronics and chips will reach 45.1 million units by 2007, with industry revenue of $1.39 billion.

Consumers today rely on all kinds of portable electronics camcorders, cell phones with built in cameras, digital cameras, laptops and PDAs. With yearly shipments of Bluetooth and WiFi devices topping tens of millions, UWB backers are eager to greet the mass market for communications and computer systems with a portfolio of products offering the kind of high bandwidth and low power capabilities that consumers have yet to appreciate but certainly will.

Roberto Aiello (roberto.aiello@staccatocommunications.com) is founder, president and CEO of Staccato Communications Inc. (San Diego).

See related chart