11n: The Same Old Stalls
May 03, 2006
Is the lack of a majority vote for the current 802.11n draft a sign of no confidence or business as usual? Depends on who you ask.
In a vote of the members of the IEEE's 802.11n Task Group yesterday, the current 1.0 draft of the high-speed wireless LAN standard failed to get the majority it needed to move on to the next step in the process. That means changes are coming, so expect at the very least a draft 2.0 (and many more drafts, if the history of the standards process is any indication).
Draft 1.0 received a 'Yes' vote from only 47 percent of the members. A draft needs a 75 percent super-majority on the letter-ballot to get approval under IEEE rules. Comments were accepted on the draft for 40 days starting March 19.
This is in stark contrast to the unanimous vote the EWC/Joint Proposal received to become the 1.0 draft, and a later 87 percent majority in March making 1.0 official. Since that time, however, numerous companies have put out chips based on the 1.0 draft, and several vendors of home networking equipment have released products using those chips. Initial tests have not gone well, indicating problems with performance and interoperability (even among products with the same chipsets).
"Consumers and business users should be wary about their purchases, at least till final ratification of the standard," according to Alan Varghese, ABI's principal analyst for wireless semiconductors.
Airgo Networks, which had strong success with products using the technology that 11n is based upon (MIMO), has repeatedly said there were problems with the draft. Their vindication with recent tests was only bolstered by the vote this week. The company released a statement: "The rejection of Draft 1.0 by the working group highlights the clear desire for significant changes... Draft 1.0 is clearly deficient in protecting the operation of installed networks."
Airgo CEO Greg Raleigh issued a statement saying that that the current vote "proves that Draft 1.0 was not as stable as some in the industry would have us believe. It calls into question the validity of so-called 'draft n' products."
The company believes the standards process was co-opted by the EWC (a consortium of competitors from Intel, Atheros, Marvell, Broadcom and others) last year (though that didn't stop Airgo from voting for the EWC proposal that became draft 1.0).
Airgo plans to make chips supporting later drafts of 11n, probably starting in the late summer of this year.
The competition at Atheros one of the companies with chips out based on draft 1.0 also released a statement saying, in effect, that this vote is no big deal: "As expected, the first ballot did not receive final confirmation. In the history of all 802.11 standards it is extremely rare for a first draft to succeed in its first letter ballot." Which is very true, but then why did companies feel the need to move so fast on products if this was so obvious? Atheros CTO Bill McFarland is quoted in the statement saying, "A 'No' vote at this stage does not indicate that there will be radical changes to the standard."
Broadcom's Director of Product Marketing, Bill Bunch, issued a similar statement saying the developments were expected and that his company "anticipates that major technical changes are highly unlikely at this point, with a 75 percent majority needed to accept any such revisions."
Atheros believes the timeframe for ratification of 802.11n has not changed, and that the final spec will arrive in 2007.