If my WLAN is G only and someone brings a B card into range of my G network, does the B card have to authenticate and associate before the G access points employ protection and I take a major hit on performance?
in order to authenticate and associate, does one not alread need to be talking at an agreed speed?
When the G network allows connections from B clients (by supporting DSSS data rates such as 1, 2, 5.5, & 11 Mbps), then when the B client authenticates and associates, the access point will announce "Use_Protection" in its beacons. Please read the whitepaper called "Protection Ripple in 802.11 WLANs" in our Learning Center for the details of exactly what happens.
A quick search failed to highlight the
We recently decided to pull this whitepaper due to some incomplete information. We came across further information that needed to be in it, and without that further information, the whitepaper presents a negative light on certain vendors. Until we can find the time to do the update, it's been pulled off the website.
As a substitute to that whitepaper, the CWAP Study Guide has an entire section on protection mechanisms and exactly how they work. thanks!
I appreciate you reminding me what the
CWNP Study book contains.
However, I'm a Planet3 CWNA and have retained a lot of the material.
The posting was a reaction to you bringing
forth the importance of the (now taboo)
Could you pull this URL from the website ?
It's still pulling up the secret report.
Thanks in advance
There's nothing taboo about it. Here are the details. Because we couldn't set up a large enough network, we made the erroneous assumption that the ripple would continue from AP to AP one after another, when in fact, it only carries to the first generation of AP. What that means is that the first AP that turns on protection will affect all APs that can hear it. APs that can hear THOSE APs will NOT be affected. Additionally, this scenario is ONLY valid for Cisco APs, not Proxim, Colubris, and others we've tested. We've now tested interoperability as well and Cisco does not cause others to use protection, but others do cause Cisco to use protection (if memory serves me right) - again, only to first generation APs from the source of "Use_Protection." WLAN Switches do not seem to have this issue for obvious reasons.
So, to sum it up, there were 2 misconceptions:
1. That it was vendor-agnostic
2. That it carried further than the first generation of APs from the source of "Use_Protection"