Aircrack-ng forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Aircrack-ng 1.7 release

Sorry Guest, you are banned from posting and sending personal messages on this forum.
This ban is not set to expire.
Pages: [1]   Go Down

Author Topic: WPA3 SAE - dragonfly handshake  (Read 17689 times)

freeroute

  • Newbie
  • *
  • Offline Offline
  • Posts: 5
WPA3 SAE - dragonfly handshake
« on: July 03, 2018, 02:49:42 am »

Hi,

Do you have any suggestion how to bypass the WPA3 SAE (dragonfly handshake)? Unfortunately with the new standard (WPA3) this tool will be useless.

"Main new features that are and aren't part of WPA3:

The dragonfly handshake (also called Simultaneous Authentication of Equals) is a mandatory part of WPA3. So if in the future you select "WPA3-Personal" in your home router, you will be using this handshake. Most importantly, this handshake is resistant against offline dictionary attacks. In contrast, personal WPA2 networks that use a weak password are vulnerable to offline dictionary attacks. Since in practice many networks use weak passwords, resistance against this attack is a major improvement.'

These is a discussion in our forum: https://forum.hashkiller.co.uk/topic-view.aspx?t=24763&m=179092#179092

Thanks for you answer.

Cheers,
Logged

tan112

  • Newbie
  • *
  • Offline Offline
  • Posts: 22
Re: WPA3 SAE - dragonfly handshake
« Reply #1 on: September 09, 2018, 09:18:12 pm »

Right, nice topic indeed.
I don't really want to spread panic between the professionals of this field of business area...
But, as I read some of the specs on that new WPA3, it seems our job is gonna be a lot more complicated...

Just take for example what happened last Friday to one of our Customers.
To make it veeery short, there was a system wireless connected, serious connecting problems between some very much proprietary-closed applications.
No matter it all looked connected right, nothing was functioning.
Out of desperation, since it was a production environment, finally we decided to try to have a look at what was actually passing on air.
So Customer authorized, we fired Wireshark with the psk password, a simple non invasive deauth, and voila, all traffic in clear.
And there we got it! That nasty "proprietary" protocol error was just passing in front of our eyes, finally!
After that it was trivial to get to the source of the real problem, and fix everything.

Now someone tell me... when that new super secure WPA3 will come around... how we, professionals, are going to solve a problem like that?
I don't really understand how those Pros at the WifiAlliance have planned this... to give just a bit more of hard time to a bunch of "hackers"... they are going to give us, professionals, a massive amount of problems... and most of problems will be without solution... and then, what I'm gonna tell to my Customer?
You know, we buy our everyday bread with our jobs, we live out of this job...

I leave that message-in-a-bottle, for anybody who has some power on that, at least to spread the facts... so some of those Gods over there might hear us miserables...
Thank you
Logged

misterx

  • Aircrack-ng Author
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1955
  • Aircrack-ng Author
    • Aircrack-ng
Re: WPA3 SAE - dragonfly handshake
« Reply #2 on: September 09, 2018, 10:55:59 pm »

The topic is mislabeled, the attack mentioned is attacking PMKID, using WPA TKIP and not WPA3.

Yes, it will be more complicated at the beginning. It was similar when WPA and WPA2 were released.

For the specific scenario you are talking about, you can always temporarily downgrade to WPA2 in the meantime. Corporate access point also have a sniffer mode that allows to capture the data and some probably can give the data as it reaches the wire.
Logged

tan112

  • Newbie
  • *
  • Offline Offline
  • Posts: 22
Re: WPA3 SAE - dragonfly handshake
« Reply #3 on: September 10, 2018, 02:41:40 am »

unbelievable ...
Unbelievable that those comments are coming from someone supposed to be a professional...
Quote
Yes, it will be more complicated at the beginning. It was similar when WPA and WPA2 were released.
False, WPA/WPA2 transition was very quiet instead, as they don't use Diffie-Hellman, and so on...
Quote
For the specific scenario you are talking about, you can always temporarily downgrade to WPA2 in the meantime. Corporate access point also have a sniffer mode that allows to capture the data and some probably can give the data as it reaches the wire.
In production environment, you can't downgrade comm protocols, or switch to wires, unless you want to lose your job...
... unbelivable
Logged

misterx

  • Aircrack-ng Author
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1955
  • Aircrack-ng Author
    • Aircrack-ng
Re: WPA3 SAE - dragonfly handshake
« Reply #4 on: September 10, 2018, 02:56:13 pm »

I'll refer to the workings of WEP compared to WPA. Assuming you know the key, it's easy to decrypt WEP. WPA on the other hand is a lot more complex to decrypt and you need key information from the handshake (temporary session) to be able to decrypt just that session. And that is the reason why a lot of tools can decrypt WEP and don't do WPA, the algorithms are more complex. WEP, you can easily construct the keystream and even do calculations by hand. WPA is a lot more complex

WPA3 uses SAE, the same stuff in mesh. It will take a bit of time before tools will be able to decrypt it.

WEP: https://en.wikipedia.org/wiki/File:Wep-crypt-alt.svg
WPA1/2 (and WPA3): https://en.wikipedia.org/wiki/Wi-Fi_Protected_Access#WPA


As of temporarily downgrading, I don't see why you would lose you job. WPA2 is still very secure, as long as your passphrase is strong. However, looking at your statement, you seem to be using PSK in an enterprise environment, which is not recommended. Not all devices in an enterprise will automatically support WPA3 when it's available and will take some time to support it. Some, such as IoT, some android or apple devices, may never support it. What I'm getting at is that you will often see a mixed environment of WPA2 and WPA3 devices and it will be very likely that such a mixed environment will be around for a few years.
I wasn't even talking about downgrading for the whole network but just for that specific device.

You can also put a tap right after the AP, where you get the Ethernet packets, which are decrypted.

My point is, there will be tools, but it might take a little while. And by tools, it also mean supplicant and authenticator.
Logged

tan112

  • Newbie
  • *
  • Offline Offline
  • Posts: 22
Re: WPA3 SAE - dragonfly handshake
« Reply #5 on: September 11, 2018, 05:35:52 pm »

Now that's a professional answer.
Thank you for the respect shown to another field professional, really appreciated.

I know my personal opinion as sole is valid nothing, but sharing it in public, I hope some other professionals might support it, if valid.
Quote
I'll refer to the workings of WEP compared to WPA
I do fully agree on that. But look at the difference between WPA3 and both WEP/WPA/WPA2 protocols.
Those protocols always integrated, now and before, a official way to decode traffic external (for whatever reasons, fallback emergencies, convenience, law enforcement investigations, and so on).
Provided of course the authorization to do so, and the most important ingredient, a strong password.
So as you say, and I fully agree
Quote
WPA2 is still very secure
It has indeed undergone few security improvements in the last years, as well as WPS, becoming now a very stable protocol in the newest hardware (and I repeat: newest hardware)

Now, look at the WPA3: it does NOT integrate any (known) way to decode external traffic, that possibility has been wiped out from the protocol foundations, that is official.
Even provided authorization, passwords, the owner looking in front of your computer, whatever, you can't decode the traffic.
Quote
My point is, there will be tools, but it might take a little while
That is true, maybe, but what "tools" are we talking about?
WPA3 officially removed the possibility to decode traffic, so the only "tools" which might come over in time will be temporary workaround bugs exploit from QA testers or worst, hackers.
As we all know very well, no respectable Customer, SME or Corporate, will never authorize the use of "hacking" tools in its environment, be it testing, stage, production or else.
(...you can see it even now, try to ask even your closer colleagues to provide a innocent VM with a cracked version of Window$... every cracked versions works just good as the original ones, yes no doubt... but just thinking of it you risk to lose your job...)

Unfortunately, that excessive "security" is going even further into obfuscating also the 80211 Open protocol.
Quote
You can also put a tap right after the AP, where you get the Ethernet packets, which are decrypted.
That you can do now indeed, and it's infact the preferred way to deal with issues in testing environments, but not in production.
Anyway, not all customers are rich Corporations willing to spend thousands of $ in such gadgets...
But as we see, with that new OWE, it will get a lot difficult even to debug a open wireless issue, bringing only frustration and loss of money, to everybody.

To conclude, my point is, security is important, but excessive "security" is destructive, and that was my evidence.
WPA3 protocol needs a fallback mechanism, OFFICIAL, in order to provide professionals (and law enforcement) a way to decode traffic in emergency situations (authorized by owners or Courts)

Thank you


Logged

misterx

  • Aircrack-ng Author
  • Administrator
  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1955
  • Aircrack-ng Author
    • Aircrack-ng
Re: WPA3 SAE - dragonfly handshake
« Reply #6 on: September 11, 2018, 07:44:00 pm »

Actually, "hacking tools" are used daily by penetration testers to check the security of networks (wired and wireless).

Regarding the Windows VM, you can provide anybody with an unactivated Windows VM AFAIK.

Taps can be fairly cheap, something like $100 and even cheaper (less than $20, search for throwing star LAN tap). Yes, you can use them in a production environment, they won't disturb anything. They've been used for IDS. Check out the options on Amazon "Network tap". The first few results are from Dualcomm. But yes, depending on the tap you want, it can get very expensive. But those are most often used for IDS purposes or large scale network capture.
Logged

tan112

  • Newbie
  • *
  • Offline Offline
  • Posts: 22
Re: WPA3 SAE - dragonfly handshake
« Reply #7 on: September 11, 2018, 08:36:40 pm »

Thank you for your reply, I'm sure someone will find it very useful. Just few remarks from my side, only for information.
Quote
"hacking tools" are used daily by penetration testers
That is not the only purpose of my job.
Braking things and getting paid for it, unfortunately is not my every day duty... mostly, I get paid to fix, and/or find the causes, of malfunctioning systems, using hacking tools is forbidden.
Quote
Regarding the Windows VM
That was just as mere example, only to describe the impossibility to use "hack" software/tools at (most of) Customer premises.
Quote
Taps can be fairly cheap
As Im' sure you have seen many times during your pentesting experience, network issues doesn't always come from the "ethernet", so a ethernet tap is most of times insufficient (besides, by most of SLAs, is forbidden to use taps in production).
Anyway, it would take ages if we would start telling about all absurdities and aberrations we have seen coming out of wireless devices.
Some of them caused by hardware issues, some by bugged firmware, some by bad quality software programmed, probably, under the effect of alcohol or other illegal drugs, and so on...
So a little piece of advice, relying solely on network taps, wont help debugging issues, as if it may come also from the wireless device...
My point is, we are clearly dealing with human errors here... hoping that stuffs will work, even after a super-secure WPA3 "certification", is just naive.
We need a fallback mechanism, to be able to revive from the hungover of that last programmer...  ;)

Anyway, I don't see this discussion is attracting much of a contributions...
Maybe that's not really an interesting topic...
Logged
Pages: [1]   Go Up