Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > RF/Microwave

Voting exposes cracks in IEEE process

Posted: 01 Aug 2006 ?? ?Print Version ?Bookmark and Share

Keywords:Loring Wirbel? IEEE? 802.20 Mobile Broadband Wireless Access Working Group? WLAN? UWB?

The IEEE's temporary suspension of activities at the 802.20 Mobile Broadband Wireless Access Working Group has unmasked festering policy problems within the broader 802 standards effort. The suspension is the first in the 802 group's 20 years of existence. But it is just the latest consequence of a long legacy of questionable practices by IEEE members that undermine the voting structure at the heart of 802's success.

Recently!and controversially!fallout from the voting issue landed on the 802.15.3a task group, which disbanded after failing to settle on a UWB approach. It also struck the 802.11n task group on next-generation WLANs, when companies launched an external consortium in a perceived attempt to muscle through a standard.

"This is a crisis. The IEEE 802 process may be broken unless the IEEE standards board can force companies to play nice," said one consultant who has attended 802 meetings since the early days of Ethernet. "And why should they need to in the aftermath of the big corporate scandals? Companies should be going out of their way to appear to play fair."

Meanwhile, progress made by such industry consortia as the Wi-Fi Alliance and WiMAX Forum suggests successful specifications can be developed outside the IEEE's working group structure.

In an interview with EE Times, IEEE standards board chairman Steve Mills and 802 chairman Paul Nikolich said the unusual 802.20 action was warranted because of the high degree of alleged irregularities within the working group.

When the suspension was announced, Qualcomm Inc.'s role in the working group took center stage, with allegations that the group's chairman, consultant Jerry Upton, had not fully disclosed his affiliation with Qualcomm. In appeals to the 802 standards board and executive committee, Intel Corp. and Motorola Inc. claim their draft proposals for the sub-3GHz licensed broadband standard were given short shrift compared with Qualcomm-based approaches.

But some independent participants in multiple 802 working groups told EE Times that Qualcomm's behavior was unusual only in the degree to which the wireless giant allegedly tried to sway the work of the committee. Other working groups, including those addressing Ethernet, WLANs and UWB personal-area networks, have likewise fallen prey to companies that try to influence the votes of participants who do not understand the issues, the sources said.

"The problem is by no means limited to Qualcomm or to the 802.20 group," said the Ethernet consultant, who requested anonymity. "Companies consistently abuse the 'one person, one vote' clause to take advantage of the process. The standards board is forced into cracking down, and if you're a legitimate independent who wants to play the game fairly, you end up getting treated like someone on the no-fly list."

Mills, however, said he sees no evidence that there are serious problems beyond 802.20, adding that the rules for independent consultants are not draconian. Nikolich acknowledged that the 802.20 group began requiring a "declaration of affiliation" several months ago and that the 802 executive committee made the same requirement at the executive level for consultant disclosures. But both he and Mills noted that most working groups do not require consultants to report all affiliations.

"The things we're seeing are only observed in a minority of cases, and only 802.20 rose to the level of necessary action," Nikolich said.

The standards board suspended all 802.20 activities until October 1, canceling a plenary meeting slated for July and interim work scheduled for September. Three appeals on the November 2005 decision to approve a draft standard!one before the standards board and two before the executive committee!are in process, Mills said. If those appeals are rejected, the 802.20 working group will pick up where it left off, at the post-draft-approval stage of early 2006, when it reconvenes.

Point, counterpoint
The 802.20 group originally considered two proposals: code-division modulation, spearheaded by Navini Networks, and flash OFDM, pitched by Flarion Inc. Qualcomm initially opposed Flarion's bid, but after acquiring Flarion in 2005, it drafted a proposal with Kyocera that incorporated some flash OFDM concepts.

The IEEE won't discuss all of the alleged voting irregularities, but confirmed that Upton has disclosed his affiliation with Qualcomm.

One former Qualcomm executive, who requested anonymity, claimed the company had brought a group of foreign nationals!specifically, a Russian team working on embedded logic for a paper mill!into the working group to vote as Qualcomm saw fit. But a Qualcomm representative called the allegations "silly," and Mills likewise doubted their credibility.

Indeed, Mills said, "no one attending a single, isolated meeting would have voting status, and we have seen no reports of continued attendance by those kinds of teams."

Two other sources who have attended 802.20 meetings, however, said there did appear to be a group associated with Qualcomm that fit the description supplied by the former Qualcomm executive.

As it confronts the appeals on the 802.20 draft, Qualcomm has presented what it claims is a strong case to the IEEE that Intel has been disrupting the 802.20 working group ever since completion of the 802.16e (mobile WiMAX) draft.

"In late 2005, active membership in 802.20!which had been around 65 or 70!suddenly ballooned by 350 percent, to more than 200, with Intel and WiMAX partners representing the bulk of new participation," said Ronny Haraldsvik, VP of broadband marketing at Qualcomm and former VP of marketing at Flarion. "We think Intel was being deliberately disruptive in order to sink 802.20."

Qualcomm will work with the IEEE standards board to find creative solutions, Haraldsvik said, but "we don't think disruptive behavior should be rewarded by removing the 802.20 draft proposal."

As for Upton's role at the working group, a Qualcomm spokesman said that several working groups had been led by active employees of Intel and other companies, and that Upton's affiliation had been a matter of public record. Most engineers who would qualify to chair groups also work for participating companies, making it difficult to guarantee neutrality, he said.

Too close for comfort?
Some members of 802.20 and 802.16e suggested the real battle is over two mobile broadband wireless standards that may be too similar in concept to coexist.

Nikolich, however, said the 802 executive committee remains convinced that 802.20 and the mobile version of 802.16 differ sufficiently to warrant separate standards. Whether Intel and Motorola, both WiMAX proponents, have entered 802.20 appeals as "spoilers" will be considered during the appeals process, he said.

Meanwhile, some independent participants in 802 working groups suggest IEEE should move to the one-company, one-vote (as opposed to one-person, one-vote) model used by many informal industry alliances, because companies are abusing the IEEE model by packing working groups.

"The motto in the 21st century should not be, 'What can you get away with?' but rather, 'What is a fair and equitable way of working with IEEE standards groups that does not reflect badly on us as a company?'," said a wireless-industry consultant. "Many bigger mixed-signal chip companies who are used to playing aggressively don't know how to turn off the chutzpah when they get into working groups. Playing like bullies seems to be the normal way to get things done."

That might have been a factor in the demise six months ago of the 802.15.3a task group, whose members had accused one another of ballot stuffing, filibustering and other delay tactics.

And when Intel, Broadcom Corp., Marvell and Atheros Communications Inc. formed the Enhanced Wireless Consortium (EWC) in October, the move was broadly perceived as an attempt to force the 802.11n task group to accept EWC's vision for next-generation WLANs. Broadcom and Atheros went on to field "draft 802.11n" Wi-Fi silicon, though offering products for draft standards is frowned upon.

Corporate pressure has even been used to create working groups on tenuous grounds. In 1995, for example, the 802.12 demand-priority LAN group was launched largely because Hewlett-Packard Co. wanted to sell its VG-AnyLAN solution.

Other working groups were created with the best of intentions and little follow-through. In 2000, Cisco Systems Inc. and Nortel Networks Inc. spearheaded the 802.17 resilient packet ring effort to replace Sonet but found few takers.

Yet Nikolich retains his faith in the working group process. While 802's executive committee acknowledges the contributions of the Wi-Fi Alliance and WiMAX Forum in developing profiles and conformance tests, the consortium organizational model is not bulletproof, Nikolich said.

"Each means of developing standards has inherent strengths and weaknesses," he said.

- Loring Wirbel
EE Times

Article Comments - Voting exposes cracks in IEEE proces...
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top