Live data from Hacker News

WiFi deauthentication attacks and home security

mjg59.dreamwidth.org

201–210 of 232 posts

Re: WiFi deauthentication attacks and home security

#202
He doesn't describe an attack, he describe literally what 802.11 was designed to do. An attack is forcing a deauth and then stealing the 4-way handshake data and, say, cracking WEP. Which is why WEP was decommissioned ... checks notes ... 15 years ago.

No need to edit aircrack-ng, WireShark does what he did natively (filter out and set channels), and a good realtek chipset allows you to set the scan interval so you can cover more channels (which is why the new ALFAs suck).

Also the DTIM and keepalive can be set such that the MCU can sleep while the phy link maintains a connection without a costly handshake, esp. if using TLS Hacking Wi-Fi has become exceptionally more difficult, as noted by the slow dating of materials at DefCon's WiFi Village over the past 8 years: cracking WPA2 is basically so hard no one bothers, even in CtF games.

Re: WiFi deauthentication attacks and home security

#203

I think the title should be changed to something like "how to protect your privacy with Wifi deauthentication". Incidentally, I was considering one of these devices as an addition to my home automation setup, then I realized that it would not be cool to monitor every person getting out of the elevators on my floor.

What if someone were to offer you, oh, $50/mo for that feed?

Do you think $50 the going rate to convince the average person to record their neighbors?

Re: WiFi deauthentication attacks and home security

#204

Earlier quoted context omitted.

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.

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 h…

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.

Re: WiFi deauthentication attacks and home security

#205
post #189

Earlier quoted context omitted.

I am not incorrect. Record audio at your peril: (This is re: New York) http://www.dmlp.org/forum/newsgathering-law/new-york-recordi... “...it is possible to violate the Wiretapping Act (and thereby commit a felony) by pointing a camera at a person speaking on a cell phone and creating an audio recording of part of the telephone conversation.” Recording audio is always fraught with risk. You should avoid it, especiall…

Isn't recording video equivalent to recording audio, as long as a potato chip bag is in view? ( https://arstechnica.com/science/2014/08/researchers-reconstr... )

It’s a hard question because of the imprecision of the law. If it’s a potential problem, talk to counsel.

Re: WiFi deauthentication attacks and home security

#206

In Norway/Oslo there is a lot of people with equipment sending deauthentication packages, jamming neighboring equipment, and one of the main reason for slow Internet (lot of jitter). Did some research on this together with The Norwegian Communications Authority (NKOM) to isolate the problem. If you want to check for yourself if someone close by i sending deauthentication packages; fire up a Mac and: 1. Open Wi-Fi-dia…

> by supporting WPA3 and PMF (encrypt management frames). OpenWRT 19.07 adds wpa3 support and the linux kernel supports 802.11w so probably many more APs could be secured.

Asus Merlin also supports PMF

Re: WiFi deauthentication attacks and home security

#207

Earlier quoted context omitted.

You might find it interesting that some German universities [1] actively send out deauthentication packages to clients that connect to SSIDs that are not on their internal whitelist to "protect" the clients from "rogue APs". A lecturer from my Hochschule was fired for protesting this practice. [1]: https://meinehochschulebehindertdaswlan.de/

From a network admin's perspective- this is necessary to protect the integrity of the air space. It discourages the use of rogue AP's which wreck the channel utilization for everyone. It's common to find this feature in enterprise wifi systems. Some actively spoof the SSID of the rogue AP in order to draw the client back to the institution's network.

From a security/technology standpoint I do not see any gain in using deauthentication packets against others. Any (half decent) skilled enough attacker will find ways to counteract these (e.g. by enabling PMF/WPA3 or using a stronger signal) or just leaving the property if a movable end device is attacked. A network and end devices are secure by design or they are not. Separation of privileges, different accounts, multi factor authentication, certificates, modern protocols, no single password policy… Making users, superiors or yourself believe that one is “secure” because of using deauthentication packets will actually result in the opposite because it will hide the existing structural security issues in your network/end devices and give a false sense of security.

What remains as argument is the service quality aspect. One wants to deliver the best wireless quality. Either because of own quality needs or because for a production environment. In either case one is using an unlicensed band that still has certain constraints based on the country. The choice of technology might just not be the right one for these needs. How about 5G? I can imagine that one can define certain rules only for a private property though. Radio waves usually do not stop on property fences. One would need to convince a court that it is 100% certain that deauthentication packets can not reach others outside of the property. I guess if the property is big enough that will work or one puts up a Faraday cage around the property or building. Still one can not send with more power as it might have health consequences for employees or visitors. Private property right is solely for how one uses the property and whom is given access to under what conditions. It does not allow a completely new rule set that conflicts with the “surrounding” law (not to mention human rights). One can still not lawfully murder a person on their private property just because the own property rules allow it.

For public universities I do not see any way to implement wireless restrictions in any lawful way. Their properties are usually public for everyone. So making rules for employees and students that can not be enforced on visitors is probably against the principle of equal treatment. Then there is the constitutional Academic Freedom in most democratic countries. Not allowing researchers or teachers to freely choose the technology suited for their needs is probably against the constitutions of these states. Also students can not be denied access to a university because of such a rule set because they have freedom of choice where and what they want to study.

Last but not least we discuss this because of a WPA design flaw that is fixed with PMF/WPA3. If we would not have had this flaw to begin with I guess we would never ever had this discussion as we do not have it for Bluetooth or wireless mice/keyboard combos or other wireless protocols that use the same frequency bands.

And finally I wonder about the mindset of the mentioned network admin’s. I can not agree if one assumes just because an organization might have more people or more important ones or richer ones or has a higher building or … to then assume might is right and enforcing it by using design flaws in network protocols. This will eventually lead to an arms race with no winner at all.

Re: WiFi deauthentication attacks and home security

#209

Earlier quoted context omitted.

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 h…

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.…

> 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.

Re: WiFi deauthentication attacks and home security

#210
post #198

Earlier quoted context omitted.

Eh? I haven't seen a single router that monitors packets on the channel other than their own, not to mention management frames of other AP's A deauth packet needs the MAC address of the AP to deauth clients connected to it and the MAC address of client you want to deauth, the latter is not required and an omission would result in the packet being treated as a "broadcast deauth" but many clients do not accept broadcas…

Just replying to point Ubiquiti APs can regularly scan channels for utilization and direct clients away from congested ones. I don’t think it has protection from deauth attempts but I think it would come across as congestion and send clients elsewhere..

Ubiquiti is hardly a common CPE, pro/enterprise grade WiFi equipment client management is a whole other story they balance based on anything from spectrum congestion to usage congestion on individual APs and the handover protocol between APs doesn’t use deauth unless it’s a very crude implementation.

There is roaming support in management frames and for signal strengths clients usually do their own roaming if you have 2 APs on different channels for the same SSID your client would select the best on and roam if necessary as the signal strength changes.

Post reply on HN