Last Post: March 16, 2005:
I am using CommView for Wifi for protocol analysis and working through trying to identify all the different frame types and exchanges and see what they look like in practice.
I have a client which I power up whilst running the protocol analyser.
I see an initial MNGT/Probe Request frame to broadcast address (what I expected). I then see a probe response from the nearby AP (also what I expected). This exchange is then repeated.
I then see another MNGT/Probe Request frame to broadcast address followed by three probe response frames.
The above repeats with various numbers of probe responses coming back to a request and then eventually we get a MNGT/AUTH frame and off we go and authenticate/associate.
I am just wondering if this is expected behaviour?
I would have expected the station to try and authenticate as soon as it received the first probe response.
Hi Andrew of UK:
I see this all the time.
A 2.4 GHz (DSSS, HR/DSSS, and ERP) client in the USA typically will send a multicast probe request on each of the eleven channels, waiting a moment after each one to receive any response(s).
A beacon producer (access point or IBSS client) that hears a probe request will send a unicast probe response, and retransmit until it gets an ACK or reaches its retransmit limit.
The less expected part comes when the transmission on, for example, channel 6 is heard with varying degrees of signal strength on channels 2 through 10, prompting probe responses from beacon producers in all those channels, and sometimes leaving a trail of retransmission attempts.
After probing all the channels the client picks the service set that is most appealing to itself based on potentially a long list of features. Fortunately each probe response carries with it the channel number of the sender.
Since a protocol analyzer can only tune to one channel at a time, and yet receives some but not all frames transmitted on nearby channels, sorting out the reason for each captured frame is a challenge in deed.
This is not so messy in the 5 GHz OFDM world where the channels barely overlap if at all.
I hope this helps. Thanks. /criss
Thanks for the response.
sorting out the reason for each captured frame is a challenge in deed.
This is the reason for attempting CWAP - new challenge needed!