3 ms·

> The most interesting one here is the deauthentication frame that access points can use to tell clients that they're no longer welcome. These can be sent for a

by oefrha 7y ago

> The most interesting one here is the deauthentication frame that access points can use to tell clients that they're no longer welcome. These can be sent for a variety of reasons, including resource exhaustion or authentication failure. And, by default, they're entirely unprotected. Anyone can inject such a frame into your network and cause clients to believe they're no longer authorised to use the network, at which point they'll have to go through a new authentication cycle - and while they're doing that, they're not able to send any other packets. Anyone has background info on why the hell WiFi spec is designed this way?

nebulous1 7y ago
The wifi alliance has a long history of failing when it comes to security audits and not doing public RFCs
- McTossOut 7y ago
  Network consortium proceedings after the 2000s have become a vehicle for companies to sabotage one another, and practically nothing else. Wifi is being used outside the much narrower scope it was probably designed for initially. Installing a cloud-based security system as your primary source of security constitutes several engineering decisions that were out of scope and it only cost $50 so short-sighted porch thieves targeted would still be plenty deterred by it.
cwingrav 7y ago
A man-in-the-middle attack is what can happen here. Deauth and then the device tries to reauth. At that point, the attacker can pose as the router and collect the password hash. The WiFi spec has serious problems.
- Someone1234 7y ago
  WiFi doesn't work the way you're claiming. You can use Deauth to be obnoxious/DoS but MITM could be accomplished without Deauth (via higher signal strength + cloned SSID) and WiFi Auth doesn't involve sending a "password hash" over the air that can be "collected." WiFi is protected via PSK (pre-shared [encryption] key), public cryptography (via CA generated key-pairs), or RADIUS. With RADIUS auth you may be able to harvest the username but the password is used as a PSK which is a shared secret between the client and RADIUS server. This is a two way check (i.e. the client confirms the RADIUS backed WiFi AP has the password too). After they both confirm each other has the password, a different encryption key is used. There's no WiFi Auth protocol that I know of that involves sending a password over the air (hashed or otherwise).
  - zamadatix 7y ago
    Absolutely false, the PTK is sent over the air and is constructed from a hash of the PMK, client/ap MAC, and client/AP Nonce. The attack the parent comment is describing is exactly why WPA3 was made with SAE. One need only capture 2 packets of the initial handshake to start offline cracking by comparing MICs and then you can decrypt the entire conversation since there was no perfect forward secrecy in WPA2 and older. Also what you describe with RADIUS is incorrect as well but there are too many ways to configure 802.1x and RADIUS to cover all of why in a comment. Overall it is considered safer than WPA2 though so the conclusion is sound.
    - Someone1234 7y ago
      > Absolutely false Let's first off go back to what I was replying to: > At that point, the attacker can pose as the router and collect the password hash. By claiming my correction is "absolutely false" you're asserting that the above statement is "absolutely true." But even your technically unsound correction doesn't actually address the underlying inaccuracy of the original statement or why you seemingly believe it is "absolutely true." It is also pretty clear from your reply that you're attempting to muddy the waters by conflating the PTK with the PSK or any other "password." The PTK isn't a password. It isn't like a password, and in order to derive it you need additional information which you need to attack (which is easier than attacking the PSK itself, thus WPA3's improvements, but doesn't make the above statement technically sound or true). Your post reads like you decided to correct before having any corrections to actually make then tried to muddy the topic as much as possible in the hope that others would be fooled. Plus is "collect the password hash" really a hill worth dying on for WiFi Auth? That's obviously an unsound technical claim, that isn't how the protocol works at all (and you seemingly must know that given your knowledge). > Also what you describe with RADIUS is incorrect as well but there are too many ways to configure 802.1x and RADIUS to cover all of why in a comment. So it is "incorrect" because I simplified it rather than describing the process in intricate technical detail? And you won't point out why it was "incorrect" because it is too technically difficult..? K.
      - zamadatix 7y ago
- oefrha 7y ago
  I’m not asking for an explanation of what the problem is (and your explanation is wrong), I’m asking about why WiFi spec is designed with this very specific, seemingly obvious flaw (anyone can fake deauth to DoS anyone else). I doubt this wasn’t considered during the design process, and I don’t think the rationale is “screw you”, so there’s gotta be a reason. Edit: According to other comments, it seems that “spoofed” deauth does have legit use cases (other than DoS’ing neighbor’s internet of shit devices).
  - zamadatix 7y ago
    The "legit" use cases people found for it were unlikely to be the reason the protocol was designed that way. Remember 802.11 came out in 1997. In 2004 WPA2 was released and allowed for protected management frames but no devices/users cared enough at the time. Now in 2019 users are more security aware and encryption is cheap so WPA3 requires it.

by oefrha 7y ago

nebulous1 7y ago

The wifi alliance has a long history of failing when it comes to security audits and not doing public RFCs

McTossOut 7y ago
Network consortium proceedings after the 2000s have become a vehicle for companies to sabotage one another, and practically nothing else. Wifi is being used outside the much narrower scope it was probably designed for initially. Installing a cloud-based security system as your primary source of security constitutes several engineering decisions that were out of scope and it only cost $50 so short-sighted porch thieves targeted would still be plenty deterred by it.

cwingrav 7y ago

A man-in-the-middle attack is what can happen here. Deauth and then the device tries to reauth. At that point, the attacker can pose as the router and collect the password hash. The WiFi spec has serious problems.

Someone1234 7y ago
WiFi doesn't work the way you're claiming. You can use Deauth to be obnoxious/DoS but MITM could be accomplished without Deauth (via higher signal strength + cloned SSID) and WiFi Auth doesn't involve sending a "password hash" over the air that can be "collected." WiFi is protected via PSK (pre-shared [encryption] key), public cryptography (via CA generated key-pairs), or RADIUS. With RADIUS auth you may be able to harvest the username but the password is used as a PSK which is a shared secret between the client and RADIUS server. This is a two way check (i.e. the client confirms the RADIUS backed WiFi AP has the password too). After they both confirm each other has the password, a different encryption key is used. There's no WiFi Auth protocol that I know of that involves sending a password over the air (hashed or otherwise).
- zamadatix 7y ago
  Absolutely false, the PTK is sent over the air and is constructed from a hash of the PMK, client/ap MAC, and client/AP Nonce. The attack the parent comment is describing is exactly why WPA3 was made with SAE. One need only capture 2 packets of the initial handshake to start offline cracking by comparing MICs and then you can decrypt the entire conversation since there was no perfect forward secrecy in WPA2 and older. Also what you describe with RADIUS is incorrect as well but there are too many ways to configure 802.1x and RADIUS to cover all of why in a comment. Overall it is considered safer than WPA2 though so the conclusion is sound.
  - Someone1234 7y ago
    > Absolutely false Let's first off go back to what I was replying to: > At that point, the attacker can pose as the router and collect the password hash. By claiming my correction is "absolutely false" you're asserting that the above statement is "absolutely true." But even your technically unsound correction doesn't actually address the underlying inaccuracy of the original statement or why you seemingly believe it is "absolutely true." It is also pretty clear from your reply that you're attempting to muddy the waters by conflating the PTK with the PSK or any other "password." The PTK isn't a password. It isn't like a password, and in order to derive it you need additional information which you need to attack (which is easier than attacking the PSK itself, thus WPA3's improvements, but doesn't make the above statement technically sound or true). Your post reads like you decided to correct before having any corrections to actually make then tried to muddy the topic as much as possible in the hope that others would be fooled. Plus is "collect the password hash" really a hill worth dying on for WiFi Auth? That's obviously an unsound technical claim, that isn't how the protocol works at all (and you seemingly must know that given your knowledge). > Also what you describe with RADIUS is incorrect as well but there are too many ways to configure 802.1x and RADIUS to cover all of why in a comment. So it is "incorrect" because I simplified it rather than describing the process in intricate technical detail? And you won't point out why it was "incorrect" because it is too technically difficult..? K.
    - zamadatix 7y ago
oefrha 7y ago
I’m not asking for an explanation of what the problem is (and your explanation is wrong), I’m asking about why WiFi spec is designed with this very specific, seemingly obvious flaw (anyone can fake deauth to DoS anyone else). I doubt this wasn’t considered during the design process, and I don’t think the rationale is “screw you”, so there’s gotta be a reason. Edit: According to other comments, it seems that “spoofed” deauth does have legit use cases (other than DoS’ing neighbor’s internet of shit devices).
- zamadatix 7y ago
  The "legit" use cases people found for it were unlikely to be the reason the protocol was designed that way. Remember 802.11 came out in 1997. In 2004 WPA2 was released and allowed for protected management frames but no devices/users cared enough at the time. Now in 2019 users are more security aware and encryption is cheap so WPA3 requires it.