The 802.11n land grab
The 802.11n wireless networking protocol has finally become a standard. It's good news for consumers but shows how slow standards bodies move.
Although it wasn't ratified until September, the 802.11n wireless networking standard has been around for quite some time. In fact, the seven-year journey to ratification officially involved more than 400 individuals ranging from equipment and silicon suppliers, service providers, systems integrators, consultant organizations, and academic institutions from more than 20 countries.
Q: What are the benefits of 802.11n?
Jay Botelho: The biggest benefit by far is more throughput--significantly more throughput--from a theoretical maximum of 54Mbps to 600Mbps with the right hardware configuration. It is this leap in throughput that makes applications like Voice over WLAN (VoWLAN) and even video over wireless feasible. It is also the reason why the claim is being made that 11n will drive more new installations to be wireless-only.
I wouldn't go so far as to say 11n is more viable than cable--each has its pros and cons. Cable (wired) handles unlimited users without effecting throughput, while wireless is shared - the more users the less throughput each one gets. On the other hand wireless is far less expensive and easier to deploy so this is a key benefit in new construction.
802.11n has been around forever it seems. Realistically, will ratification translate to a surge in deployments?
Botelho: Many enterprises held off with 11n upgrades (and therefore wireless upgrades in general) for fear that the ratified spec would be substantially different from the Draft2.0 spec (the one the Wi-Fi Alliance based its pre-ratification certification tests on). Now that this question is no longer an issue, and since there's probably some pent-up demand since wireless upgrades in general may have been put on hold, it is expected that there will be a surge in deployments. As an aside, there is very little difference between the Draft2.0 version and the ratified version.
Give me a sense of the upgrade process from older standards.
Botelho: I think this will be the bulk of the upgrades. People will upgrade (a) for the bandwidth and (b) because a/b/g (previous standards) hardware will just get harder and harder to purchase. I think what we're seeing with 11n is a real transition from somewhat ad-hoc deployments--a little wireless here and a little wireless there--to well-planned and well-executed rollouts of wireless as the primary network access technology.
Such a transition may require wired upgrades as well, since 11n gear really needs to be connected to a gig switch and not just a 10/100 switch (another reason why many have held off until now). I also think the managed wireless LAN market will be in its heyday as most customers choose a complete, managed wireless LAN solution over DIY deployments of gear from the likes of CDW.
What about upgrades from preratified 802.11n deployments? Any considerations?
Botelho: I really don't see much activity here. Most preratification 11n gear, with the exception of those predating Draft2.0 and the Wi-Fi Alliance (WFA) precertification testing, will work just fine in the post-ratification era. In fact, gear that was certified by the WFA against Draft2.0 is being grandfathered in against the ratified standard--the changes in mandatory features are really that insignificant.
What do you expect will be some of the challenges 802.11n will have to overcome in 2010?
Botelho: I see challenges with wireless management and the convergence of wired/wireless management. Most wireless monitoring and management tools available today were designed with the needs of a b/g network in mind. As stated above, with a 10x increase in throughput the amount of data to analyze goes up by the same factor. In many cases, this is a re-architecture of the software.
Users [IT staff] will also be looking for combined wired/wireless management systems. This has driven some recent acquisition activity, like Fluke purchasing AirMagnet. At this point, combined wired/wireless management systems are not ready for prime-time, and it's unclear if they ever will be truly integrated. The two systems are just different, and require different views into the data and different types of fault detection.
One other possible challenge is the overall development of 11n as a technology. Up until now, most systems have been "2x2" systems, meaning they employ two data streams on the transmit side and two on the receive side. 11n can support up to "4x4" systems, and it is only at this level that the theoretical maximums of 600Mbps are met. Though a great deal of time and testing have gone into the 11n specification to ensure compatibility, I remain a bit skeptical as to how all this will roll out, especially when it comes to interoperability.