[Madwifi-devel] Tests on 802.11h with madwifi_trunk_r4100

[Dennis B. <den...@go...>]

Hello madwifi-list!

I did some testing on the 802.11h stuff, that I got from the trunk 
version of madwifi. I took the last version available, r4100. My setup 
was the following: I had one Alix Board, that was monitoring everything, 
that was going on on channel 124, the other Alix-Board was the AP and 
always used channel 124 on bootup(SSID="master_00000001") ( I configured 
this via a script, that did the wlanconfig-stuff and afterwards ran a 
`iwconfig ath0 channel 124`). My observations are the following:

If I bootup the system and just sniff what's happening on the wireless 
medium, there is quietness for 60 seconds (as wanted by the 
regulations), but then, there are only 10 beacons to come and after that 
there is nothing anymore (see tcpdump-file "no_traffic").

If I bootup the system with the same settings and afterwards start a 
software, that produces multicast traffic, the traffic gets transmitted 
the way it sould be, but there are no beacons to see (see file 
"multicast_traffic").

If I bootup the system and at the same time I have another AP running on 
the same channel(SSID="test"), the system behaves exactly the same way. 
It first sends 10 beacons and as well transmits the multicast messages. 
So although there is traffic on the wireless medium, the AP transmits 
the frames and does not care.

So all that I observe, is an AP, that boots up, stays quiet for 60 
seconds, then emits only 10 beacons, stops transmitting beacons for no 
obvious reason but does not care about traffic on the device at all - 
neither does it care for another transmitter being on the same net, nor 
does it stop sending multicast messages (which I would expect, if it 
stops transmitting beacons).

So my question in general - is there something, that I misconfigured or 
that I do not understand, or is this a bug in the 802.11h-code of 
madwifi-trunk_r4100?

Kind regards,
Dennis