Last week, the Wi-Fi Alliance unexpectedly announced it would reverse is oft-repeated decision to do no interoperability testing of 100+Mbit/s wireless LAN gear until final, formal ratification of the 802.11n standard. Instead, the industry group now plans to start testing in mid-2007 wireless LAN chipsets, components, NICs, and access points based on an emerging draft of the standard. Various vendors and industry analysts are predicting that the IEEE 11n task group will publish a 2.0 draft standard by early in 2007. Network World talked with Alliance managing director Frank Hanzlik about this sudden change in the Alliance's position.
What did the WFA actually decide to do?
We decided to create a two-phase certification program for 802.11n. The first phase will be to certify products based on the IEEE's progress in the first half of next year.
The second phase will be certification concurrent with the final ratification of the standard, which is expected to be in March 2008.
The Alliance repeatedly insisted it should and would wait for a final standard before trying to test and certify an expected flood of "pre-standard" wireless products. Why the reversal?
First, when we set up our 11n roadmap a couple of years ago, our position was we would wait until the 11n standard was ratified. But we expected then that ratification would be in March 2007. But now, it's more likely to be March 2008, a year later than our original operating assumption.
The second point is there are a lot of pre-standard products now shipping into the marketplace. And there will be a lot more in 2007. ABI Research has some recent data, which predicts tens of millions of pre-standard 11n products shipping in 2007.
If we don't have a standard in place, and a certification program, the market will be confused, and it won't come together because there won't be industry-wide interoperability.
Those are the two considerations that drove us to develop this two-phase plan.
So was this an occasion for acrimonious debate by the board of directors?
I wouldn't use that term. This is certainly something we've been looking at for a while. There are clearly a variety of different perspectives on what makes sense [for us to do]. We worked toward a full consensus. We believed this was the right path forward.
We continue to remain aligned with the IEEE on its longer-term direction [for the standard]. And yet we're showing industry leadership with [going ahead] on interoperability testing in the shorter term.
We did something not too dissimilar with security when we created Wi-Fi Protected Access (WPA) and WPA2 [which were based on a not-yet final version of the IEEE 802.11i security standard]. They were very well received by the marketplace. That experience led us to think that the same kind of model could be used in the case of 802.11n as well.
Is the Alliance membership united on this decision?
This strategy has very, very broad support across our membership. I think you're going to see a very, very large amount of product coming out when we roll out this test program. With 11g, we were just crushed by the number of products [submitted for testing]. We're working hard to make sure that problem doesn't' happen again.
When the final standard is ratified [in 2008], and we bring out the final certification program, one goal is to have backward compatibility with the pre-11n equipment that will come out next year.
What does this decision mean in practical terms?
We hope to have, first, an approved IEEE letter ballot [of draft 2] by March 2007. That looks to be a realistic expectation. We think a 2.0 draft will be put out in the January meeting for ballot. We'd know if that was accepted or not by March.
We'll encourage progress in the IEEE on that ballot. That will be an important milestone.
And if that doesn't happen?
We believe it's still important for us to launch a certification program. We'll take the latest version of the draft [at that point] and design our certification [testing] program around that.
Are you working on the test program now?
We set up some 11n task groups in December 2005, one working on the marketing side to define the requirements, one working on the technical side [of designing the tests].
We've already had a couple of 'plug fests' with early [pre-11n] products, to get more experience with the underlying technology and to help us figure out which elements to incorporate in a robust test plan.
When will testing begin?
We're saying June 2007. It could be earlier, but at this point, I think it's better to be a little more conservative.
Are there any special requirements for testing high-throughput wireless LAN products?
It's safe to say that it's the most advanced 802.11 technology we've had to test in the last 6 years. In recognition of that, as I mentioned before, we started our process earlier than we would normally have done, to build up our experience until we're in a position to roll this out to the market.
Do you own any pre-11n wireless equipment?
I personally have not bought any pre-N products. Mainly because what I have is working fine for me now.