Detecting Spoofed MAC Addresses On 802.11 Nets
Joshua Wright writes "I have written a white paper on detecting spoofed MAC addresses on wireless LAN's. This paper describes some of the techniques attackers utilize to disrupt wireless networks through MAC address spoofing, demonstrated with captured traffic that was generated by the AirJack, FakeAP and Wellenreiter tools. Utilizing the techniques I describe, it is possible to identify users who utilize spoofed MAC addresses on 802.11 networks to launch denial of service attacks, bypass access control mechanisms, or falsely advertise services to wireless clients."
From the amount of the whitepaper I skimmed through, it looks like this could be used over copper, but the type of attack that it detects is rare or non-existant on copper because of the inherent difficulty of taking down a single client with DoS without taking down the entire network itself.
No, it will not apply to copper. The packet generation in 802.11x includes a counter. This counter is not present in the 100/10tx packets. The counter is generated at the physical (hardware level), and so when an intruder attempts to DoS a valid user and usurp the mac address, the counter cannot be changed to what the user's counter was...
UNLESS...........
the intruder either waits until the user's counter is about to flip back to 0, then DoS the user, and reset his counter, then spoof the MAC address. Or perhaps a virus or trojan could be written that would reset the valid user's counter somehow.
Basically what this guy did was realize that the MAC-generation algorithm in spoofing software Wellenreiter has a weakness, namely that the OUI's it generates aren't all legit. (OUI is the organizational unique identified which is in the first few bits of the MAC address.) Also see helpful Sourceforge description of Wellenreiter.
He similarly points out limitations in denial of service tools: AirJack and FakeAP software. However, this isn't the same as giving a general technique for analyzing MAC addresses on 802.11b, something which was strongly implied in the original post.
http://tinyurl.com/4ny52