I'm looking for a solution to implement a rogue AP policy that classes all rogue APs broadcasting an SSID containing non-ASCII characters as malicious, in order to avoid phishing attacks such as the one described in the following article: https://air...