IEEE STANDARDS PROCESS ON UWB TO VET NEW TEST RESULTS

At the center of the debate in an IEEE working group is work toward a single UWB standard for wireless personal area networks, which has left a split between major equipment- makers. One draft proposal for the 802.15.3a standard is backed by the Multiband-OFDM (Orthogonal Frequency Domain Modulation) Alliance (MBOA), which includes Intel, Nokia, Panasonic, Samsung, TI and others. The MBOA proposal would rely on a frequency-hopping technology over multiple bands, which supporters said was allowed under the FCC’s Feb. 2002 order on UWB. A CDMA-based proposal developed by Motorola and XtremeSpectrum, which Motorola agreed to acquire last month, is based on a “traditional” UWB system that uses a broad area of spectrum, also under limits set in the FCC order. The standard under development would cover short- range UWB applications, allowing data rates of 110 Mbps at 1 m and up to 480 Mbps at shorter ranges.

The MBOA proposal at an IEEE working group meeting in Albuquerque, N.M., last month again failed to win a 75% confirmation vote, leaving in play both that plan and one backed by the Motorola camp. The inability of MBOA to win the threshold vote was the latest in a series of working group meetings that failed to reach closure on the issue. The battle over those standards has been hard-fought, with several observers likening them to the outcome in the Wi-Fi arena over 802.11b. In the meantime, technology developers on both sides of the debate said they were moving forward in the early stages of product development rather than waiting for the standards process to sort itself out.

“I think it’s going to continue to be in a stalemate,” said Martin Rofheart, dir.-UWB operations in Motorola’s Wireless & Mobility Systems Group. But he said the standards process itself was “only made more relevant by waiting.” He said were the MBOA solution to be ratified by the IEEE working group now, the approved solution would be based on power point presentations and not chip sets. He said new technologies typically went through a preregulatory development phase, which UWB now was past with the Feb. 2002 adoption of the FCC’s order. That means UWB is in the 2nd stage of “vertical markets served by proprietary solutions,” he said, which entails the commercialization efforts of launching a first set of products into a series of devices. The market then vets these solutions from a perspective of what makes sense on performance and cost, he said. “It is only after those things have been completed that we can enter a stage of a meaningful, ubiquitous standard,” he said. “Any standards body that adopts a solution before it is vetted by industry runs the risk of being irrelevant.” He said Motorola’s UWB group was ahead of MBOA members on the development curve because the company already offered a chipset.

Disagreements persist between the Motorola camp, which also includes MBOA, Oki Semiconductor & Communications Research Lab and ParthusCeva about the interference potential of the proposed standard backed by the Intel-TI alliance. At issue is whether the kind of frequency-hopping (FH) system backed by MBOA would have to be tested for compliance with the hopping function turned off. Motorola has said that depending on whether its or MBOA’s interpretation of testing requirements is correct, FH devices may emit energy far in excess of FCC limits, in spite of passing a compliance test, or they might have to operate at lower power levels than non- FH systems in order to comply. MBOA members have disagreed with those arguments.

Testing has indicated that when FH systems are turned on, the signal wave form is “more interfering than the FCC actually allowed,” Rofheart said. “I'll be very interested to see what testing they have done and what the results are.” For an FH system to meet power limits with the hopping function turned off while the transmitter is on, devices would have to reduce their transmit power substantially, Rofheart said.

On the interference issue, Steve Turner, UWB business development manager for TI, said the information on the test results, to be presented at the IEEE working group’s meeting in Vancouver, B.C., next month, should address concerns by MBOA opponents. “We will have some information to show what we have been saying all along, which is that there isn’t a problem,” he said.

Intel’s Wood said the experimental test results confirmed simulations MBOA members have conducted on interference issues. “We did some limited experimental testing against some specific devices that could be argued as being interference targets,” Wood said. “That testing corroborated the paperwork that we did. We are in the process of engaging the FCC to see if we have done enough testing and if they would like us to do a little more.”

Aside from interference concerns, another sticking point in the standards debate has involved intellectual property rights. XtremeSpectrum, just before it was acquired by Motorola this fall, filed a 2nd “letter of assurance” with IEEE indicating it would offer zero-licensing access to patents involving its proposal for the 802.15.3a standard. TI said earlier this month it would offer royalty-free licenses for its UWB technology based on IEEE approval of the MBOA standard. Fellow MBOA member Wisair made a similar announcement this month.

TI’s Turner said MBOA members ultimately still would like to see a standard blessed by IEEE, but he said: “The market can’t afford to wait.” A 0.9 draft of MBOA’s specification is set to be released around Feb., with a Version 1.0 slated to be out around May, he said. The goal is to work on silicon development in 2004, with samples by the 4th quarter, he said. “We are hopeful that the situation in the IEEE is resolved by then,” he said.

Apart from the IEEE standards process, an ITU Study Group that begin meeting in Jan. has been working on UWB policy issues, including definitions and operational and technical characteristics. The next meeting of the group is set for June in Boston.

Some European administrations had raised concerns at an ITU task group meeting this fall in Geneva about the potential impact of UWB on IMT-2000 systems, or the European version of 3G wireless, several sources said. A contribution by a working group, which is examining the compatibility between UWB devices and wireless services, outlined results from several studies. In part, the document evaluated the impact on an IMT-2000 handset when UWB devices were at varying distances. “The mobile is completely swamped by the UWB device when it is within 0.5 m,” said the paper, which compiled different study results to compare them side by side. At a distance of 1 m, the amount of intracell interference that an IMT-2000 receiver can tolerate is reduced by 10% of what it would be when there was no UWB device, the paper said. It said more studies were needed to evaluate in greater detail the impact of certain UWB power spectral density limits on the IMT-2000 network. The paper said UWB interference always had been considered to be “white noise” with UWB device deployments depicted as relatively isolated, “always-on” transmitters. But the paper said some study results showed a correlation between the distance of UWB and IMT-2000 devices, “which emphasizes a potential for greater interference power into the IMT-2000 networks.” The paper said other studies were based on UWB devices operating at far less frequency than “always-on” systems, which it said could reduce the impact of interference into IMT-2000 networks.

Another contribution to the task group examined the compatibility between the Europeans’ DVB-T (Terrestrial Digital TV) and UWB systems in the VHF/UHF bands. Among the scenarios examined was one in which the DVB-T receiver antenna was located outdoors and an interfering UWB transmitter operating in both an outdoor and indoor environment. The potential for interference involves more than one UWB transmitter, the paper said. “This time the multiple interference is due to the professional use of UWB LAN, for example in buildings having glass walls and large open-space work places (with negligible indoor-to-outdoor loss), which would potentially generate high aggregate interference to DVB-T receivers operating nearby,” the paper said. Using FCC emission limits that were adopted in Feb. 2002 and UWB slope emission masks, the paper said the results of the analyses examined “clearly show that the FCC UWB emission limits do not guarantee the protection of the DVB-T system in the presence of UWB emissions, while the UWB slope emission masks reduce significantly the interference probability. Consequently, the UWB slope emission mask concept should be adopted at European and international levels to protect the DVB-T system in the presence of UWB emissions, in the VHF/UHF bands.” The study focused on UWB systems operating in the 3.1-10.6 GHz bands to DVB-T systems operating in VHF/UHF spectrum.

Several sources said many of the same interests who raised objections during the FCC and NTIA consideration of UWB commercial applications had been raising concerns in the ITU arena, including cellphone makers and service providers and groups concerned about potential interference to GPS and other satellite-based systems. “The same parties that opposed ultra-wideband in the U.S. are opposing it internationally,” a source said. The source ticked off land mobile, satellite and fixed microwave interests, as well as radioastronomers. “The land mobile people are major players in Europe, in particular,” the source said. The source said UWB policy discussions at the ITU level might not reach the same kind of resolution as had been achieved in the U.S., in which the FCC and NTIA brokered an agreement to allow rules to go forward and to reconcile competing technical claims about interference. Because the ITU is consensus-based, “what comes out of the ITU is a bunch of recommendations, not rules,” the source said. “Individual countries decide how they want to take those recommendations. Frankly, it’s hard to see how this is going to reach the same conclusion [as the FCC] on the international level.”