Could the 802.11n Logjam Finally Be Broken?
Last Friday, in a vote even some veteran observers of the networking industry weren't expecting, members of the IEEE 802.11 Working Group voted unanimously -- 100-0, with five abstentions -- to advance what's currently being called Draft 1.1 of the 802.11n high-speed WiFi standard. A subsequent vote in the spring could move the draft to 2.0 status, even though as recently as last November, the ratification of Draft 1.06 left behind, by one official count, 370 outstanding catalogued technical issues for further discussion.
With so many issues on the table, what's expected to be the final ratification of a Draft 3.0 standard, in which those issues are resolved, is tentatively scheduled for October 2008, according to a report in InfoWorld. The WiFi industry can't wait that long, and is apparently allowing last week's vote to serve as the starting gun for a new wave of "Draft-N" equipment.
But which draft are the industry players talking about? Last month, on the same day it announced its acquisition of MIMO technology leader Airgo, Qualcomm announced it was sampling the first chipset to offer "full support for Draft 2.0 of the IEEE 802.11n standard." Even now, there is no Draft 2.0, although last week's IEEE vote made Draft 1.1 look a lot more like a 2.0 in the wings.
If Qualcomm is the "sooner" in this land rush, Intel may be the "boomer." Yesterday, Intel announced it's producing a "Next-Gen Wireless-N" Mini Card adapter, as part of its evolving Centrino Duo platform. But small print in Intel's technical specifications brochure for the new card, model 4965AGN, defines "Draft-N" as referring to the 1.0 standard - the one voted down overwhelmingly by the Working Group last May. Intel has been part of a coalition, along with Broadcom, Marvell, and Atheros, to pursue as far back as late 2005 the production of chips and products that claim to follow the 1.0 draft.
The key difference between the old and new drafts has been how newer 11n equipment will coexist with older 11g and 11b equipment within the same coverage areas. For 11n's optimum throughput of 100 Mbps to be possible, it requires a 40 MHz channel. Under the rejected 1.0 draft, that channel overlapped with spectrum already allocated to 11g and the other standards, with the idea that everybody's most likely going to be upgrading anyway, so if it overlaps, that'll give consumers more incentive to trade up. That idea didn't fly.
The solution proposed in the new 1.1 draft, and that will likely be carried over into 2.0, is a re-allocation of 11n's spectrum to two 20 MHz channels. Using the radio signals WiFi already uses to determine whether a transmitter is in the area, an 11n device will detect a working 11g device in range first, and if it's active, the 11n device will temporarily shut off its second 20 MHz channel.
But here is what may be the remaining point of contention: Qualcomm's "Draft-N" devices currently being sampled will reportedly use MIMO, the multiple antenna technology it acquired from Airgo, to boost the speed on that remaining 20 MHz channel, even if the first one is shut off when an older device is in range.
While Draft 1.0 utilized a minimum of two antennas as well, Airgo chose to pursue its own MIMO implementation, which Qualcomm soon backed - and not too soon thereafter, bought. Networking equipment which doesn't really follow the letter of Draft 1.1, but may follow the spirit of it -- or some part of it -- may not adopt this option.
Since last August, the Broadcom coalition has opted to go on a different track, which veteran industry observer Glenn Fleischman calls "a slightly parallel universe." Literally moments after last week's Working Group vote, Broadcom announced the availability of its Intensi-Fi Wireless LAN chipset. But while Broadcom applauded the vote, it stated its new product "incorporates all mandatory elements of the IEEE 802.11n draft specification and is designed to be software upgradable once the standard is finalized."
Broadcom did not say "2.0," "1.1," or even "1.0." However, Intel confirmed late this afternoon to BetaNews that its Next-Gen Wireless-N products will try to follow the 1.1 draft. "We are compliant with 1.1 and expect to be with 2.0," the spokesperson told us. "The materials were written and tests done when it was just 1.0, we will be updating our materials and disclaimers as this rolls along." Broadcom has also been contacted for clarification, which may be forthcoming.
As long-time writers in the networking field have pointedly observed in recent months, it isn't exactly technically feasible for a company's products to be "compliant" with a draft of a standard. Prior to formal publication, drafts by design have unresolved issues, and Draft 1.1 of 802.11n is certainly no exception.
But others feel that the contention issues between different wideband WiFi players will literally be played out physically - not in the market, but in actual everyday use. And by the time the IEEE gets around to publishing the final 802.11n specification, you and I may have already solved this problem...albeit with some bumps and bruises along the way.
3:20 pm January 24, 2007 - A spokesperson for Broadcom told BetaNews this afternoon that it believes all the Intensi-Fi products it has been shipping since last year are in compliance with the most recent draft of the 802.11n specification, though the spokesperson declined to attach a draft number.
"We are very happy to confirm that all of our chipsets shipped to date are compliant with the most recent IEEE 802.11n draft specification," the spokesperson wrote, "and we are confident that our Intensi-fi products will be upgradeable to the final standard."