Reality game
Long ago, the process of setting technology standards was the work of noble engineers who wanted to create technology for the greater good. While engineers who currently develop such standards remain pure in heart and mind, their employers make the game fundamentally different because those technology companies that play their cards right in the standards game stand to control entire markets. In short, standards development has become a multi-billion dollar industry.
“Standards are fundamentally political and have nothing to do with technology,” said Carl Cargill, chief standards officer for Sun Microsystems, who is a 20-year veteran of the standards-development process. “Standards are fundamentally a social issue. A finite number of them have a tremendous impact on the economy, and people are using standards to capture that market. That’s not what they were intended for.”
Case in point: The Institute of Electrical and Electronics Engineers, or IEEE, took an unusual step in June by halting work on the 802.20 standard — which is designed to compete with the 802.16e Mobile WiMAX standard — after Intel and Motorola threatened to file formal complaints about the way the working group’s chairman, Jerry Upton, allegedly handled draft proposals in favor of Qualcomm and Kyocera.
Upton called himself an independent consultant, but reportedly revealed to the IEEE that he had a relationship with Qualcomm. Intel and Motorola had been complaining since mid-November 2005 that Qualcomm, which subsequently merged parts of its technology submission with Kyocera’s proposal, submitted an incomplete technology proposal and violated other IEEE procedures.
It’s clear that 802.20 technology will be a competitive threat to WiMAX, a technology heavily favored by Intel and Motorola. The 802.20 standardization process had been spinning its wheels for more than two years, but contributions from Qualcomm revived the process late last year. Qualcomm was working to use its considerable influence on the 802.20 working group to finalize a standard by the end of 2006, in a bid to provide an alternative to Mobile WiMAX.
Qualcomm dominated the standards process as it held an estimated two-thirds to three-quarters of the votes in the 802.20 working group because the IEEE counts votes from individuals rather than on a per-company basis. However, it was the company’s alleged abuse of the process that resulted in the IEEE restarting the 802.20 standardization effort in September, after appointing new officers and establishing new balloting and balloting-resolution committees. It also tightened the requirement for disclosure of affiliations.
“It was pretty flagrant, and it was the right thing to do,” Brian Kiernan, head of the standards unit with technology company InterDigital and member of the 802.20 standards working group, said of the IEEE’s move. “Anyone can abuse the process, regardless of what type of process it is.”
‘EVERYONE KNOWS THE GAME’
Working the standardization process has become an art form — especially on a global basis. The Chinese, who see technology leadership as a way to gain significant influence in the world, are learning fast. In June, the Chinese delegation walked out of the opening of a two-day meeting in the Czech Republic that was scheduled to discuss harmonizing WAPI — or wireless local area network authentication and privacy infrastructure, China’s homegrown encryption technology for Wi-Fi networks — with international standards from the International Standards Organization (ISO).
The move followed Chinese accusations that the IEEE used underhanded tactics to prevent the global approval of WAPI. The IEEE has chosen 802.11i as its encryption standard. The Standardization Administration of China, in a statement, accused backers of the American technology of “a lot of dirty tricks, including deception, misinformation, confusion and reckless charging to lobby against WAPI,” the Chinese newspaper Xinhua reported.
Cargill recalls a conversation he had with one of the Chinese delegates regarding the situation. The gist of the conversation was that the Chinese saw this loss as only the first in a series of battles. “That sent shivers down my spine,” Cargill said. “They have nothing to lose by organizing.”
It appears the Chinese are rallying several third-world countries in Africa and South America for votes in the ISO. Earlier this year, China and the U.S. each submitted their technology choices for Wi-Fi encryption to the ISO for standardization. According to ISO rules, national bodies voting against either of the two technologies should attend the meeting and explain their votes. But only seven of the 17 countries opposing WAPI showed up at the meeting. Recently, the Chinese hosted a summit where every African nation participated. It did the same with South American countries.
Despite the intrigue and controversy regarding WAPI, it’s clear that many members of all standards bodies, including the IEEE and the Telecommunications Industry Association (TIA), do have a purist attitude when it comes to making standards. Despite the drastic measures IEEE took to revamp the 802.20 standards-setting process, Steve Mills, the IEEE’s Standards Board chair, said the vast majority of participants in IEEE standards activities have “always acted in the good-faith and cooperative spirit that has resulted in high-quality and broad adoption of IEEE standards.”
“Everyone knows the game,” Kiernan said. “The engineering guys know the game … and those issues are there, but they don’t try and decide those issues based on [intellectual property rights]. They do try and come up with the best solution.”
For example, the vocoder adopted by the APCO Project 25 committee is the Improved Multi-Band Excitation, or IMBE, vocoder developed by Digital Voice Systems. It was chosen through an objective listening process to understand which vocoder had the best quality, said Larry Nyberg, standards strategy manager with Motorola. “We lost the vocoder battle but subsequently switched to IMBE and retrofitted quite a bit of radios,” he said.
But there remains a fine line, especially when IEEE standards are voted on by individuals, not by companies. “The reality is, anyone who goes to the IEEE is funded by a company,” Cargill said. “We send a lot of people to the IEEE, and we pay for them. Are they individuals? Hell no. We send them in with distinct instructions about what position they are to take. The IEEE recognizes this.”
In fact, the IEEE, in an effort to make the standards-setting process more transparent, announced a significant change to its patent policy last month, much to the chagrin of some of its largest members. Effective April 30, the IEEE is promoting advance disclosure of ambiguous or restrictive licensing caveats associated with patents that might be included in its standards.
“Until now, we’ve allowed the inclusion of essential patents in IEEE standards if patent holders assure us they will license their patents without compensation or with reasonable and nondiscriminatory terms,” said Judith Gorman, IEEE managing director. “However, the lack of specific information in such assurances may create uncertainty that can impede the adoption of a standard.
“The new policy is an important step in correcting this situation through additional transparency of licensing terms for the technology alternatives being considered for inclusion in a standard,” she said. “The policy should also benefit anyone who seeks to comply with IEEE standards after they are approved.”
In particular, the voluntary policy permits and encourages the optional and unilateral disclosure of royalty rates and other licensing terms before a patented technology is included in the standard. The disclosed terms may include, for instance, the maximum royalty rate that the patent holder will look to charge.
That means a lot to a marketer. If one piece of technology costs $2 versus 1¢ for a slightly less optimal one, the marketer is going to take the 1¢ option, Cargill noted. But the new patent policy also is threatening to those players who customarily cross license, so it’s unlikely some large technology companies will want to voluntarily offer up their rates.
HERDING CATS
With this political backdrop, the creation of standards is, needless to say, an arduous matter complicated by the fact that market forces generally determine how fast a standard moves along, and a lack of competition can leave a standard stagnant.
For instance, the Project 25 standard, first offered by Motorola, is more than 16 years in the making. Competitors say that sloth-like pace is because Motorola was the dominant player in the public-safety wireless market, and there was little impetus for the standard to move along for the first 10 years, when few competitors were looking at interoperability.
“Motorola was the sole proponent of P25, and the only larger system provider of P25, so there wasn’t much emphasis or point in making a standard because the company was locked into a sole-source position,” said Paul May, business development manager for M/A-COM. “M/A-COM began selling and deploying systems, and then there were two separate systems houses with fundamentally the same system. All of a sudden, there was an apparent need for a standard.”
Nyberg counter-argues that competitors were trying to keep a market going for analog equipment for as long as they could.
At any rate, it’s clear that meaningful competition in a given market is a clear driver for standards development. The broadband standard WiMAX, for instance, has enjoyed the backing of several major vendors — including Intel, Motorola, Nortel and Samsung — and has moved rather smoothly (if that’s possible in the standards world) through the development process. WiMAX also has benefited from a strong commercial marketing entity, the WiMAX Forum, which is dedicated to certifying compatibility and interoperability of WiMAX products and support standards.
Meanwhile, P25 until recently has consisted primarily of an over-the-air standard and the promise of many more to come. The standard’s inter-radio frequency subsystem interface was finalized last summer. But to be fair, public-safety communications in the U.S. are not found in a single swath of airwaves like WiMAX. Instead, such communications are scattered in pockets throughout the spectrum chart, which makes standards development more challenging. Also, P25 attempts to address the needs of thousands of small public-safety agencies that may not be able to buy a new system for 15 to 20 years — instead of developing a standard for greenfield deployments by well-funded entities — so technological backward-compatibility is a must, no matter the complications.
While market forces can significantly slow the progress of a standard, the process itself is laborious. Typically, about 100 engineers — from market competitors representing divergent business and patent interests — must agree on a single standard in the end.
It’s akin to herding cats.
“A standard by definition is not the best technical answer. It’s what everyone will agree to,” InterDigital’s Kiernan said. “And that’s not the absolute best technical approach.”
It takes a special breed to play the standards game. “Engineers in TIA and IEEE aren’t kids out of school,” Motorola’s Nyberg said. “They have a lot of ownership in what they are doing. It takes a very good communicator, a thick skin and patience because sometimes it’s worse than watching paint dry.”
MANY GAMES TO PLAY
Proposing a new standard is the easy part. Getting other potential stakeholders interested is the hard part. “It’s easy for companies to block a proposed standard by joining their technological resources to try and find holes or use parliamentary rules to block the adoption of a standard,” May said. “There are many games you can play.”
At a basic level, both the IEEE and TIA call for a negotiation process that sets the requirements for the standard, a process that often takes longer than the entire development effort itself, Kiernan said. Once that is accomplished, engineers begin the proposal stage that will knock around various approaches that are submitted by either a single company or a consortium of them. These discussions are governed by Department of Justice rules, which don’t allow participants to talk about marketing, selling prices or equipment costs playing a role in a standard.
More often than not, two or more factions develop that force standards bodies push for the two sides to compromise. In an extreme case earlier this year, the IEEE task force considering ultrawideband proposals overwhelmingly voted to disband the standards process after failing to reach a consensus on a single IEEE 802.15.3a standard. The fight between the UWB Forum and the WiMedia Alliance over standards had been raging since 2002.
It’s also not uncommon for much of the heavy lifting concerning a standard’s specification to come out of business forums — such as the WiMAX Forum, the CDMA Development Group or Bluetooth Special Interest Group — because all of the interested vendors are usually members of these forums. However, these groups aren’t accredited standards groups and don’t have the anti-trust exemptions groups such as the IEEE and TIA have.
Once proposals are nailed down, standards bodies such as the IEEE and TIA attempt to ratify them using various approaches. Some groups use a formal ballot process while others pass proposals if no one objects.
“Ultimately, when you get a ballot [to approve a standard], there are people sitting in meetings for two years who do not say a word, but if they vote ‘no’ on a ballot, as long as they have a technical objection, TIA rules say we have to bend over backward for those objections,” Nyberg said. Some TIA documents have taken two years to reach a technical consensus in the balloting stage.
The IEEE balloting process involves two steps. There is first a working group ballot and then an IEEE sponsor ballot. Most IEEE members agree that although this is a long process, it results in a stronger standard.
According to Kiernan, who served as a vice chairman for the WiMAX 802.16e standard, the 802.16e working group rewrote the document six times and the first ballot vote generated 2000 technical comments, all of which had to be addressed. It took one year for the standard to pass the IEEE sponsor ballot. Both the IEEE and TIA require a 75% consensus.
One major difference between TIA and IEEE is TIA’s one-company, one-vote policy versus IEEE’s individual voting policy, although IEEE can collapse the vote and require companies to vote instead, as was the case with the 802.20 standard. The one-company, one-vote process isn’t foolproof either, as companies can in essence “buy” each other’s vote as they compromise.
The sticky issue of patent declaration certainly plays a key part in the standards process. As mentioned earlier, IEEE is trying to offer more transparency. TIA requires companies to identify what they hold, but it isn’t very explicit about the point in the process at which companies must reveal their patent holdings, Nyberg said. However, companies can’t allow a standard to be published and then later declare they hold intellectual property rights (IPRs) for the standard. A number of court cases have established a precedent in that regard.
The P25 process has some unique IPR stipulations. Early in the process, a memorandum of agreement was created that requires companies to tell the steering committee what their intentions are in terms of licensing, as well as disclosing the basic licensing terms. That lets any company entering the P25 market compare notes, so to speak, to ensure the licensor is staying true to what it promised TIA.
Oddly, despite the contentiousness, these engineers — who are spending years battling it out — can have a drink together. As Kiernan put it: “In general, most people respect each other and their opinions. They can be partners today and enemies tomorrow.”
THE MAKING OF 802.20
OCTOBER 2005
-
Qualcomm submits its proposal for 802.20 specification.
NOVEMBER 2005
-
Intel engineer complains that Qualcomm’s proposalis incomplete.
JANUARY 2006
-
IEEE narrows 802.20 standard consideration to joint proposal from Kyocera and Qualcomm.
-
Intel engineer files statement asserting that the working group included 20 or more consultants that appeared to be engaging in improper voting as a block.
MARCH 2006
-
Qualcomm/Kyocera proposal receives majority of votes.
MAY 2006
-
In another vote, Qualcomm/Kyocera proposal receives majority of votes.
-
In appeals to the 802 standards board and executive committee, Intel and Motorola claim their draft. proposals for the 802.20 standard were given short shrift compared with Qualcomm-based approaches.
JUNE 2006
-
Intel and Motorola threatened to file formal complaints about the way the working group’s chairman, Jerry Upton, allegedly handled draft proposals in favor of Qualcomm and Kyocera.
-
802.20 Chairman Jerry Upton discloses he is a consultant for Qualcomm.
-
IEEE postpones 802.20 standards work to look into allegations from Intel and Motorola.
SEPTEMBER 2006
-
IEEE announces reorganization of working group that included the appointment of new officers as well as new balloting and balloting-resolution committees. It also tightened the requirements for disclosure of affiliations.