Slashdot Mirror
With WPA3, Wi-Fi Security is About To Get a Lot Tougher (zdnet.com)
One of the biggest potential security vulnerabilities -- public Wi-Fi -- may soon get its fix. From a report: The Wi-Fi Alliance, an industry body made up of device makers including Apple, Microsoft, and Qualcomm, announced Monday its next-generation wireless network security standard, WPA3. The standard will replace WPA2, a near-two decades-old security protocol that's built in to protect almost every wireless device today -- including phones, laptops, and the Internet of Things.
One of the key improvements in WPA3 will aim to solve a common security problem: open Wi-Fi networks. Seen in coffee shops and airports, open Wi-Fi networks are convenient but unencrypted, allowing anyone on the same network to intercept data sent from other devices. WPA3 employs individualized data encryption, which scramble the connection between each device on the network and the router, ensuring secrets are kept safe and sites that you visit haven't been manipulated. Further reading: WPA3 WiFi Standard Announced After Researchers KRACKed WPA2 Three Months Ago
One of the key improvements in WPA3 will aim to solve a common security problem: open Wi-Fi networks. Seen in coffee shops and airports, open Wi-Fi networks are convenient but unencrypted, allowing anyone on the same network to intercept data sent from other devices. WPA3 employs individualized data encryption, which scramble the connection between each device on the network and the router, ensuring secrets are kept safe and sites that you visit haven't been manipulated. Further reading: WPA3 WiFi Standard Announced After Researchers KRACKed WPA2 Three Months Ago
--dave
[English, ambiguity is your middle name]
davecb@spamcop.net
Yes, this will prevent open-air sniffing of your packets.
VPN or HTTPS is still better, because after those packets arrive at the access point, they are unencrypted over whatever wire the AP is plugged into. WPA only covers the wireless link; HTTPS or VPN (or both!) encrypt much farther through the network, if not the whole way.
The first thing I do on an open WiFi network is connect to a VPN.
Slashdot still doesnâ(TM)t support Unicode after it was added to the HTML standard in 1997.
"One of the key improvements in WPA3 will aim to solve a common security problem: open Wi-Fi networks. Seen in coffee shops and airports, open Wi-Fi networks are convenient but unencrypted, allowing anyone on the same network to intercept data sent from other devices. WPA3 employs individualized data encryption, which scramble the connection between each device on the network and the router, ensuring secrets are kept safe and sites that you visit haven't been manipulated"
Sure. But your computer will still not know that the CoffeeShop SSID that they're connecting to was the one the shop set up, though, will they? There's no exclusivity for SSIDs and if there was, it'd be a denial-of-service opportunity.
Once connected, and a secret shared, yes. But with no password the initial connection is still giving people a chance to shove you on THEIR connection rather than the one you think, and then you can be WPA3-authenticated to them rather than what you thought without having a clue.
The WPA to end all WPAs.
There needs to also be some kind of certificate system added for open networks. Starbucks ought to be able to register their network with a CA, so that itâ(TM)s possible to verify that that open network with the SSID âoeStarbucksâ is not a phishing network.
I believe that in some countries like Germany it is illegal to run an open wireless network. (Crazy but true!) Would this proposed new standard address that, since the network would now be encrypted and no longer 'open'? Or does the law define an open network as one where users don't have to register for a username first? In that case, open Wifi would sadly remain illegal in Germany.
-- Ed Avis ed@membled.com
In the coffee shop scenario WPAx is often not even active. Unless WPA3 also works for connections without a passkey, how does this help?
We should be on WPA4 or 5 by now or moved on to another 3 letter security like WTF.
I wonder what caused the 13 year wait?
Minimum threshold fixed. Thanks!
FBI Chief Calls Unbreakable Encryption 'Urgent Public Safety Issue'
Right.
I'd love to see something like WiMax come back with open support so anyone can run something with longer range. It sucks that wifi has such a short range, but LTE can go so far, costing you a fortune per gig. It would be nice to have something that anyone can setup that covers longer distance, even if it's at a reduced speed.
> a "password in a file" would be the private key, but even that isn't really a good comparison, because you never transmit your private key
Since at least the 1980s (Kerberos) and dial-up modems used CHAP in 1996, you can authenticate via a password without transmitting the password.
There are even better algorithms that use passwords, without transmitting or storing them on the server. For example, the server can store a salted bcrypt of the password. Upon login, the server generates a random number (the challenge) and sends that to the client, along with the salt the server has chosen for this user. The client then computes and sends:
H(H(Hs(password, salt)), challenge) xor Hs(password, salt)
The server can verify that without having the password transmitted, or stored on the server.
You would be correct to say that *sending plaintext passwords over the network (1970s style)* is much less secure than public keys. You can certainly use passwords without sending them over the network, though - that issue has been solved for decades.
> Plus, even shitty private keys (1024 bits) are way stronger, entropy-wise, than a password so there's that, too.
Much like a LONG password (pass sentence).
Can't you just turn off file sharing, and "discovery", or whatever it's called? I mean, unless somebody gets into the router, you shouldn't been seen, right?
Backport for the WRT54GL when?
The article said Tougher, not Better.
PKI infrastructure required for every home wifi or Windows 15 will not connect. That's pretty tough.
That said, are people sure than "unsecured" WiFi is not encrypted today? Would fail against man-in-the-middle but not against evesdropping.
What happens when the password is written on the wall of the coffee shop? Can anyone with that password break encryption for others? Can anyone with that password be a man-in-the-middle?
Anything of any import, even damned cat videos, are secured by https these days. If someone sniffs your packets, all they see is cyphertext
The ClientHello message that opens a TLS session contains the destination hostname in cleartext, so that the server can tell which name-based virtual host's certificate to present.
Since at least the 1980s (Kerberos) and dial-up modems used CHAP in 1996, you can authenticate via a password without transmitting the password.
Yes, true. (Although it's 2018 and I have yet to see an ISP that wouldn't use PAP)
There are even better algorithms that use passwords, without transmitting or storing them on the server. For example, the server can store a salted bcrypt of the password. Upon login, the server generates a random number (the challenge) and sends that to the client, along with the salt the server has chosen for this user. The client then computes and sends:
H(H(Hs(password, salt)), challenge) xor Hs(password, salt)
The server can verify that without having the password transmitted, or stored on the server.
Interesting, although the last authentication protocol I've heard to have that property (MS-CHAP, not that I knew many authentication protocols) was broken.
You would be correct to say that *sending plaintext passwords over the network (1970s style)* is much less secure than public keys.
Yes, however it's not like "sending plaintext passwords over the network" wasn't a common thing in 2018.
You can certainly use passwords without sending them over the network, though - that issue has been solved for decades.
Yes. You've just explained one and referred to another. You got your point across, no need to be redundant.
> Plus, even shitty private keys (1024 bits) are way stronger, entropy-wise, than a password so there's that, too.
Much like a LONG password (pass sentence).
Well, who uses a LONG pass sentence? It has to be pretty long if you limit yourself to actual words and want to compete with 1024bits of random data, which means 128 bytes from an alphabet the size of 256.
And it's not even like I would deny that the private key is like a password in a file.
The statement I was refuting is that the *certificate* is like a password in a file, because it's clearly not, and that point still stands.
CLI paste? paste.pr0.tips!
I'm always surprised Kerberos didn't get used for WPA.
Radius servers with WPA2-Enterprise is interestingly far more secure than the WPA2 most people use at home.
- Michael T. Babcock (Yes, I blog)
Check out www.spriteguard.comthose guys are dedicated to leveraging security obligations onto providers and not joe public. Very, very innovative stuff!