* 802.11n has many performance dials and knobs
As noted last time, new types of potentially attractive, useful 802.11n products are set to emerge that the Wi-Fi Alliance is currently not equipped to test and certify.
The looming arrival of 1×1 (single-stream) 11n products for small, power-challenged devices and even for simply more affordable laptops is throwing the Wi-Fi Alliance’s operational chart into a tizzy. I should note that the performance-certification issue in the 802.11n market is far broader than forthcoming 1×1 products, though they are most likely to force the issue.
I’ve written, from time to time, about the need for an objective performance-certification lab where 802.11n products could be compared fairly. 802.11a/b/g were founded on pretty straightforward technologies that delivered a consistent performance range. There were fewer optional knobs to turn and enhancements to capitalize on to make products from respective vendors operate at dramatically different rates.
For those products, it makes sense to be able to go to the Wi-Fi Alliance Web site and look under the type of device you want (laptop, network adapter, access point, wireless router) to see which ones have been 802.11a- or 802.11g-certified for interoperability.
802.11n is a different animal. Depending on implementation (number of spatial streams supported, whether optional features like beamforming are supported and so forth), performance can vary widely. As you likely know, the Wi-Fi Alliance conducts interoperability tests for all Wi-Fi products but makes no claims about product performance. For political reasons, it’s unclear if that’s even a job the alliance should take on, given that funding is provided by the Wi-Fi vendors whose equipment the alliance tests. Can you imagine the arguments over the testing protocols, specs and conditions? It could be a task argued over ad infinitum and never completed.
I know that for about the past year, Wi-Fi testing company Veriwave has been working with technical services company TUVRheinland in Germany to see about creating an enterprise-backed, objective lab for 11n product performance certifications. As you might suspect, funding has been the sticking point. Enterprises do seem interested in third-party performance-based certifications. But they’re reluctant to fund it. Indeed, which user companies should pay while others don’t? (If you can think of a possible model, please feel free to discuss it in the community or send me an e-mail.)
In the interim, it has been suggested that different categories of Wi-Fi Alliance testing for 802.11n products become instituted based not on the type of device, but on the number of spatial streams supported. In effect, these would be certifications that map to “performance categories” rather than to device types. In this way, the problem of a 1×1 laptop not being testable and certifiable because it doesn’t meet the Alliance’s definition of 802.11n for laptops (2 spatial streams required) would go away.
For example, products could be certified in tiers, labeled something like 802.11n-1, 802.11n-2 and 802.11n-3 indicating spatial stream support. Or they could use monikers like 802.11n-100, 802.11n-200 and 802.11n-300, indicating the rough throughput expected with each tier.The Wi-Fi Alliance acknowledges the need for changes. Marketing director Kelly Davis-Felner says a working group has been scoping requirements and that the Market Requirements Document is mature but not yet approved.
“It’s safe to say we are not ignoring the issue,” says Davis-Felner. “We’re definitely aware of feature variations and the potential for problems with end user confusion. We’re looking at it with a lot of carefulness.”
The Wi-Fi Alliance plans to address the issue by the time the final 802.11n standard is ratified in 2010, she says. However, the impact of 1×1 devices that launch this year remains unclear.
