Hello Jeff, Did you get something out?
I read quite a lot through the official docs but did not get something clear. Actually when I study
the relation between client scannings and the snr reported by APs, I have following findings.
My client device was performing continuous scans back to back (probe request again right after receive responses of the previous request). Suppose the client heard from APx (may or may not be the associated one) at T0, and (by pulling out snr statistics from the controller using snmp query) further suppose the controller has a new snr record about the same client from APx at T. Then I found T - T0 on average (across 50+ APs) is 300+ seconds...
Does it make sense? Note that the client may not have any data traffic with these APs...
Regards,