I noted in a recent column that I thought the Wi-Fi Alliance would produce an interoperability specification based on a draft of the 802.11n standard. The alliance took this step last week. The move is unprecedented, controversial, not entirely clear in terms of scope or long-term impact -- and absolutely the right thing to do.
Just to review, Wi-Fi is a term often used synonymously with "wireless LAN." But Wi-Fi is a set of interoperability specifications that enables products from different manufacturers to work together. It does not, I must note again, specify compliance with the 802.11 standard. Its relationship to 802.11, while certainly more than incidental, isn't hard and fast. The alliance was not formed to promote 802.11. Rather, as a trade association, the alliance was formed to look out for the interests of its members, promoting common interests among companies that are otherwise cunning competitors. It can, in short, do whatever it wants.
The long delay in the development of the 802.11n next-generation standard forced the alliance to reconsider how to best fulfill its mission. While the alliance has issued interim specifications in the past (most notably WPA, in an effort to correct the deficiencies in WEP), it has never undertaken the certification of something as large as a new physical layer so far in advance of the completion of work by the IEEE. In many ways, the alliance's members forced the organization's hand. Again, developing an interim specification is a good move because it provides a path to correct one of the most colossal marketing errors in the history of WLANs.
What so many WLAN vendors have done to create this problem is now well-established. They decided to come out with lines of "Draft n" products well in advance of the IEEE actually issuing the 802.11n standard. Many companies assumed that their Draft n products would ultimately be upgradable to the full standard. But with .11n now unlikely to appear before mid-2008, such promises seem doubtful at best. The lack of interoperability among the different Draft n products compounded the problem. Top off all of this with the fact that earlier products based on multiple input/multiple output (MIMO) provide much better performance (not to mention interoperability) than the Draft n units, and the need for action was quite clear. The decision by the alliance to issue an interim interoperability specification is the best course of action now.
The plan is to do exactly that in the second quarter of next year, basing this spec on the current state of the .11n draft (presumably 2.0) at that time. While it may be impossible at that point to guarantee whether products can be upgraded to the standard (let's see if vendors offer such a guarantee then), at least we'll have a foundation for interoperability from an association that knows how to do just that.
The implications for the current crop of Draft n products are now clear -- they should be positioned as a bridge to the past, a "better g than g" if you will, and ultimately terminal in and of themselves. I'm expecting that products based on the upcoming Wi-Fi spec will offer much better performance, and we'll see the beginnings of a battle for both pure performance (throughput vs. range) and price/performance that will continue right through the introduction of products based on the approved standard likely in the second half of 2008. My guess is that big-time price cuts will shortly be the order of the day on the current crop of Draft n products, which, by the way, still offer much better throughput than plain old .11g.
Note, though, that the alliance's interim specification will not be 802.11n. I'd urge the alliance to avoid mentioning anything about "n" in this spec other than its stated goal that the spec that it issues for the approved standard will be backward-compatible with the interim spec. Let's leave it up to the vendors to promise (or not promise) forward-compatibility and avoid opportunities for more confusion.
One important implication for the enterprise is that the alliance's action should result in a good number of enterprise-class MIMO products next year. While some companies will choose to wait for products based on the approved standard, many will begin to deploy MIMO to take advantage of the much-improved throughput only available via this technology. And I think the enterprise WLAN vendors will offer cheap (if not free) upgrades via the swapping of a mini-Peripheral Component Interconnect or similar module.
In the meantime, the residential WLAN market will likely go all-MIMO next year, a big boost for the vendors and exactly what a trade association like the Wi-Fi Alliance always has as its primary goal.
Craig J. Mathias is a principal analyst at Farpoint Group, an advisory firm specializing in wireless networking and mobile computing. He can be reached at craig@farpointgroup.com.