WPA Weak Key Cracker Posted

Glenn Fleishman writes "The folks at TinyPEAP released a cracking tool to break Wi-Fi Protected Access (WPA) keys. WPA is the replacement for weak WEP keys in the original 802.11b specification. Robert Moskowitz of ICSA Labs released a paper almost exactly a year ago documenting how WPA keys that were short and lacked randomnness could be subject to cracks. This tool automates the process. Moskowitz advised choosing passphrases of more than 20 characters or generating random keys of at least 96 bits, but preferably 128 bits. Some tools exist to produce better keys, including chipmaker Broadcom's SecureEZSetup (in selected hardware) and Buffalo Technologies' hardware-based AOSS for automatic key generation and propagation. Enterprise-based WPA with 802.1X doesn't have this weakness: each user gets a long WPA key that's randomly generated and uniquely assigned--and can be frequently changed during a session."

Comments

By its nature...

[The Islamic Fundamen]

When you really think about it, by nature wireless networking can never be too secure. I mean, your data is being broadcasted across the air to another point. Think about it.

Re:By its nature...

[Excen]

This is true, but if it's encrypted it might be considered to be marginally safe from tinkerers like your average /.er. That being said, stay away from my wireless network you damn kids!

Re:By its nature...

[davesplace1]

You make a good point, I know that I would not do any "online" banking with wifi.

Re:By its nature...

[Anonymous Coward]

When you really think about it, by nature the internet can never be too secure. I mean, your data is being transmitted through dozens of other servers to another point. Think about it.

Re:By its nature...

[wcdw]

Theoretically, perhaps - but how secure does it need to be? All wireless traffic in my home uses SSH tunnels between the laptop and the firewall.

When it becomes possible to conveniently crack SSH tunnels, I'll start to worry. By then, I'm sure there will be something better available. Meanwhile, you can sniff those ESP packets to your heart's content.

This is trivial under Linux, and not much more difficult under Winblows (clients), and I'm surprised more people don't suggest it as an alternative to WEP/WPA.

(My girlfriend uses Winblows w/ SSH Sentinel, and has only had one problem that rebooting wouldn't fix - in over 3 years. That one? Installing XP / SP2 turns on the [useless] firewall, which blocks the ports needed by the VPN.)

http://www.theboyz.biz/Computers, parts, electronics, small appliances and more!

Re:By its nature...

[slashdot.org]

When you really think about it, by nature wireless networking can never be too secure. I mean, your data is being broadcasted across the air to another point. Think about it.

I guess that's an understandable misconception about security. But security has by nature nothing to do with wireless or wired.

Good security is based on the principle that other people WILL have access to your encrypted data.

Unfortunately, the people that implemented security in the wireless protocols did a piss-poor job and left it vulnerable to (known!) attacks.

However, if you just ran IPSec or something over your wireless connection, you'd be fine.

Re:By its nature...

[Anonymous Coward]

Installing XP / SP2 turns on the [useless] firewall, which blocks the ports needed by the VPN.)

That's what firewalls do...they block ports. Be they SP2 or some other variation this has nothing to do with the design of Microsoft's firewall and everything to do with the concept of firewalls.

But since you've called Windows "Winblows" and labelled te firewall useless it's obvious you're lacking sufficient knowledge to know why the problem occured.

Re:By its nature...

[Anonymous Coward]

Who modded this insightful? Of course proper security always assumes that everything you transmit is captured inbetween.

Real security makes it really hard to use the captured data without the key, which should never be transmitted cleartext.

Re:By its nature...

[wcdw]

The FACTs are that when SP2 was installed, it altered the system configuration, and installed a perfectly useless product. (Actively dangerous, as noted by the bug which enables file/printer sharing across ALL connections if you have it on any!)

As for not knowing what happened, it took me about 10 seconds to solve the problem. And, in fact, DID require a reboot, but then again, that's Winblows.

As for lacking sufficient knowledge of firewalls, you're welcome to try and hack mine. It's been up for 7 years now without an intrusion. And not for trying, according to my logs.

Re:By its nature...

[Alejo]

SS1 is weak in many ways. are you SSH2 only?
Also, how good is your w32 software on picking session keys? host keys?
Not saying you're insecure, just that you didn't mention basic 101 stuff about it. And ssh implementations on w32 are rumored to be all weak. :-/

Re:By its nature...

[Lord Kano]

Do you have an alternative link for SSH Sentinel? ssh.com no longer has it available.

LK

Re:By its nature...

[wcdw]

Actually that's a finger slip (trying to hard to remember the name of the SSH Sentinel product ;) - actual implementation is IPSec VPNs. However, I do have all my SSH clients and servers set up for SSH2 only, and use SSH/SCP exclusively for remote access. Even across the VPN. ;)

Re:By its nature...

[KillerCow]

When you really think about it, by nature wireless networking can never be too secure. I mean, your data is being broadcasted across the air to another point. Think about it.

Your wired network can't be too secure either. All that you need to do is attach a listening device to a wire somewhere. Or just compromise a machine.

See the sibling post about how the basis of cryptography is asuming that someone has access to your encrypted data and the encryption algorithm. All security rests in the key. Cryptographic algorithms exist that can make it infeasable to decrypt a block of cyphertext without the key.

Re:By its nature...

[wcdw]

Interesting; I didn't realize they'd dropped that product. I didn't find a copy locally, but the original download is probably still on my girlfriend's laptop - although I'd have to check redistribution licensing.

It's a shame; it makes it MUCH easier to do IPSec than the built-in XP VPN feature. (Their connector product looks comparable, but I really don't know anything about it.)

Re:By its nature...

[KingPunk]

just generate a key from /dev/urandom on nix. doesn't get any more random than that.

and im fairly certian it won't be compromised any time in the near future
;)
--kingpunk

Re:By its nature...

[Fweeky]

Looked at OpenVPN? Seems a lot easier to configure than a VPN.

Re:By its nature...

[Fweeky]

Er, s/a VPN/IPSec/. *cough*

Re:By its nature...

[AngusSF]

Try ftp://ftp.up.ac.za/pub/linux/ssh/pub/sentinel/

Re:By its nature...

[drfrogsplat]

As for lacking sufficient knowledge of firewalls, you're welcome to try and hack mine. It's been up for 7 years now without an intrusion. And not for trying, according to my logs.

Put your money where your mouth is and post your IP on /. then (;

Re:By its nature...

[Anonymous Coward]

why does an inflatable doll need access to the internet?

Re:By its nature...

[Anonymous Coward]

Wireless is obviously less secure than wired networks... but these are also vulnerable to:
- TEMPEST attacks (measuring the electromagnetic field around the wires to find out what information is going trhough)
- unprotected switches allowing an external laptop to plug into the company network (and to arp-spoof you)

To be secure, you sometimes have to be paranoied... for instance, I take for granted that all the packets on my network can be seen, either by a war-driver or by someone using the wired network.

The only response is: encryption.
You want to secure the trafic between your wireless laptops and your servers? Use IPSEC! Better yet, L2TP/IPSEC (compatible with the MS VPN client).
You want to provide secure services? Force the use of HTTPS (when relevant) and POPS/IMAPS. Use the SASL/TLS capabilities of your MTA (who said postfix?).

And always remember: the level of security for a network shall always be directly linked with the sensitivity of the data going trough.

Re:By its nature...

[Jacco de Leeuw]

There are alternatives.

Re:By its nature...

[Alejo]

still missing: session keys and host keys. :)
Plus how good is your OS at getting entropy? What symmetric encryption algorithm? What key exchange algorithm?
And about ssh over vpn... a friend (known player @ crypto) told me once that you should never assume that re-encrypting would improve security, unless you are using a well known and tested method of mixing both encryption systems.
Intention of this post is not bitching, but to try to make ppl aware it's not just "i use XXX, so im safe", but a very complex subject.

Re:By its nature...

[Tim C]

You do yourself and your arguments no favours when you call it "Winblows"; it just makes you look petty and immature. Same goes for people referring to "Lunix" and "open sauce" or "open sores".

Re:By its nature...

[RandomJoe]

It is a WHOLE lot easier! Reading the discussions, I was wondering if anyone else had comments on it. I was originally trying to set up IPSec for home, but had the dual problem of figuring out how to get my work (Win2K) laptop using it (while not messing up the VPN client my company had set up), and just plain figuring IPSec out in the first place. What a mess... I could get there, but next time I needed to do it (very seldom) I was learning all over again. (Yeah, take notes, I'm bad about that...)

I then tried OpenVPN, and without much difficulty at all have set up connections for both wireless access at home, and remote from work to the house, on both my Linux laptop and the work Win2K laptop. The connections on the work laptop were set up probably 2 months after the Linux one, and it only took me a few minutes to remember how to do it. (Using RSA keys, not preshared keys.)

I'm no security expert, so I have to rely on what is said on the OpenVPN site and elsewhere. Is this pretty trustworthy? I now have it setup so NOTHING happens over wireless unless you VPN somewhere. Either OpenVPN to my home network, or the work laptop can VPN (Cisco client) to the corporate office. Remote access is the same way, and limited to certain IPs that I'm likely to be at.

Having done this, I also don't bother with WEP/WPA, but do put the MACs in the AP. Yes, they can be spoofed, but then they hit a blank, unresponsive wall, except for the OpenVPN port. My firewall is not "standards compliant" - I just DROP undesired packets from WLAN or Inet. Fun to see those "test your IP" sites asking if I'm sure the computer is on! ;-)

Re:By its nature...

[mrph]

It is, in fact, a VPN in the same way as other VPN's.
Still, you're right. It is really easy to configure, has clients for various OS's.
I've been using it for about a year, linking my home LAN to a remote network (two computers act as
VPN gateways which means it is all transparent for the systems using the connection) and I really like it!

Re:By its nature...

[Anonymous Coward]

Theoretically, perhaps - but how secure does it need to be? All wireless traffic in my home uses SSH tunnels between the laptop and the firewall.

The best security is layered. If you used WEP/WPA they would have to break both.

But this isn't the biggest misconception you have. Your computers likely run other services.. uPnP, ICMP and others. One that watches what goes out not just what comes in.

And if your connected to the internet others could use your services for nafarious uses and not even look at your computers. And if you have mac filtering on your "ifconfig ath0 ether 00:aa:de:ad:be:ef"

But your right about the SP2 firewall, it us useless. Get a free version of Zone Alarm, or better yet Zone Alarm Pro or some other commercial firewall.

Re:By its nature...

[wcdw]

Speaking of misconceptions, I know _exactly_ what services all of my machines are running. In fact, even if you were to break into one of my machines at the keyboard, you still would have highly limited access to the other machines.

Why in ANYONE's name would I be running nPnP?

In case anyone is wondering, I've done SysAdmin work in real environments since the early 90s, and spent many years prior to that breaking mainframe/mini/micro security from the programmer side. Securing a couple modest home networks is hardly a full-time challenge.

Although what that has to do with end-to-end encryption being easier than WEP/WPA and safer than MAC address filtering... <shrug>

Re:By its nature...

[wcdw]

Yes, there are people who say that. My take on it is that anyone who can't get over my calling it Winblows has some petty and immaturity issues of their own with which to deal...

And heck, I never even once typed Windoze instead.

Re:By its nature...

[wcdw]

What - you think you can hack my firewall, but you can't even find the IP from which I'm posting? Sorry, no cardboard cookies for you!

Re:By its nature...

[SillyNickName4me]

> just generate a key from /dev/urandom on nix. doesn't get any more random than that

Maybe you are talking about a specific implementation here (Linux I bet) and detaisl are sightly different between different unix like systems... /dev/urandom is not random at all, it is pseudo-random at best.

The basic issue is that as soon as you think up a process that generates numbers in a way that you can describe mathematically, you also end up with a process uncapable of generating real randomness.

You can get most aspects of randomness, but what you won't get, and that is the most important part for encryption, is unpredictability.

How predictable things are depends for a bit on the algorithm that you use, and for a large part on the abbility to deduct the current state of the 'random generator'. If those 2 are known, the next number your random generator will produce can also be known.

This is why it is so important to have a good entropy source, it makes it virtually impossible to guess at the state of the generator.

Re:By its nature...

[wcdw]

All good points, from a security point of view. I should point out that I do stay current on security patches, including randomness issues, and that ALL of the connections in these networks involve at least one Linux box.

Regarding SSH over VPN, I don't do it for added security, and am familiar with at least some of the dangers multiple encryption layers can present. I do it because when the laptop is wireless, it CAN'T talk to anything without the VPN -- and there is no command line access to any of my boxes save SSH, even through the wired network.

And actually, the nature of the root post should make it clear that 'just because I use [WPA] I'm safe' is a fallacy. It's entirely possible to make SSH and/or IPSec relatively insecure.

For example, using a pre-shared key that is the name of your dog, and e-mailing it to the receipient on the other end, well...

But as far as making people aware, as far as I know, no one has even been able to get across the concept of strong passwords, never mind creating memorable ones that don't need to be written down anywhere.

It's unlikely that people are going to change, meaning that encryption needs to be make stronger IN SPITE OF the user. It wouldn't be that hard, for example, to add a routine to new WPA boxes that refused to accept weak passwords.

Re:By its nature...

[Fweeky]

Yes; in my defense it was 4:11am when I wrote that, and I spotted it as soon as I hit Submit. Meh ;)

The main thing holding me back atm is the lack of a PocketPC or Zaurus client (the latter's easier to deal with, but don't mistake me for someone who isn't lazy); IPSec would be better in that regard, if I could be bothered learning how to set it up.

Enterprise mode (Radius-authenticated) WPA seems interesting too, but every Radius server I've seen seems hideously overcomplex for just a little home WLAN.

Re:By its nature...

[Anonymous Coward]

Apologies if this question has been discussed before, but how does someone set this up (IPSec over wireless) from the wireless card to the router? I have a Win2000 box so I can't use WPA (without paying money I mean).

Re:By its nature...

[TheLink]

The big difference is current popular wireless networks are a lot easier to jam than current popular wired networks.

And that's good enough reason to not use it for anything important.

Re:By its nature...

[peter]

> /dev/urandom is not random at all, it is pseudo-random at best.

On Linux, that's wrong. /dev/urandom returns very high quality pseudo-random at _worst_. /dev/random never resorts to mere pseudo randomness, and read(2)s on it block until the kernel has accumulated enough entropy in its pool. (yes, Linux maintains an entropy pool which it seeds from random events so there is some true randomness waiting for programs like gnupg or statistical simulations that need it.)

You're correct about everything else, though. The only thing you didn't know is that /dev/random doesn't come from a purely algorithmic source. Kernels have access to more than just a Turing machine :).

> This is why it is so important to have a good entropy source, it makes it
> virtually impossible to guess at the state of the generator.

Now you're talking. That's why Linux uses the low bits of the CPU's clock cycle counter sampled during interrupts (which are generated by disks, the network, keyboards, and mice, etc. i.e. fairly unpredictable things, esp. wrt. exact numbers of CPU cycles!) It mixes these samples into its pool with cryptographically strong algorithms (insert hand-waving here... :), so even if the samples aren't very random, they don't make it worse.

If you're totally paranoid, RML's netdev-random patch will let you choose whether you want to add entropy from network interrupts to the entropy pool. Of course, you could also use rngd from rng-tools to feed entropy from your chipset's built-in rng (which measure thermal noise, and so has randomness that fairly directly from quantum mechanical processes, the only known source of true unpredictability in the Universe.)

Re:By its nature...

[SillyNickName4me]

> On Linux, that's wrong. /dev/urandom returns very high quality pseudo-random at _worst_. /dev/random never resorts to mere pseudo randomness, and read(2)s on it block until the kernel has accumulated enough entropy in its pool. (yes, Linux maintains an entropy pool which it seeds from random events so there is some true randomness waiting for programs like gnupg or statistical simulations that need it.)

Blahblahblah.

1. the point of my post was to point out that you should verify that your random generator has a good enough entropy source, and if you had bothered to read my post a bit more carefully, you would have seen that I am aware of the fact that Linux does a decent job at this.

2. You ensure randomness in the entropy pool, and thereby in the state of the random generator. The generator itself however is still pseudo random.

3. Sorry if I sound annoyed here, but what was the point of your post other then trying to push a specific system?

Re:By its nature...

[peter]

> 2. You ensure randomness in the entropy pool, and thereby in the state of the
> random generator. The generator itself however is still pseudo random.

I think a pseudo-random generator continuously re-seeded with true randomness will produce truly random output, not pseudo-random output. Running true randomness through a good mixing function shouldn't destroy it, and neither should taking hashes of parts of the entropy pool.

Cryptographic strength is all about predictability, but for simulations, true random numbers matter because you don't want to skew things with any possible patterns.

> 3. Sorry if I sound annoyed here, but what was the point of your post other then
> trying to push a specific system?

You said the data you get from /dev/random is pseudo-random. The point of my post was that it's truly random because it comes from truly random events, like the precise timing of a keypress. I used Linux as an example because I know how it works, not purely to push it. I guess I can see how you might have gotten that impression, though, so no hard feelings.

Re:By its nature...

[SillyNickName4me]

> I think a pseudo-random generator continuously re-seeded with true randomness will produce truly random output, not pseudo-random output. Running true randomness through a good mixing function shouldn't destroy it, and neither should taking hashes of parts of the entropy pool.

It depends on a limited number of potentially observable events, all of which interact in a known way. This doesn't make for real randomness.

If this could be used in any practical way for predicting the outcome of the random generator is another question (it is at least extremely difficult I think)

> You said the data you get from /dev/random is pseudo-random. The point of my post was that it's truly random because it comes from truly random events,

See above. If all events would be observable, it is possible to know the state of the random generator. If this is practical depends on the situation, but I can think of situations where it is practical.

There is also the situation that some or many of those events may not occur or will occur in a known patern.

Last but not least, it is still relevant if what you do also results in a well distributed use of all states of the generator over time, else you severely limit the actual entropy that you achieve.

Don't get me wrong btw, what it does is good enough for many uses, but it isn't truely random.

Is this a bit of hair splitting over a definition? well, not entirely. It is good to know that it is good enoug, for now, yet not perfect by any means, and may not be usable for all applications right now either. (think embedded devices that need strogn security and have very little 'random' events to use for this, or where restricting physical access is impossible)

> Cryptographic strength is all about predictability, but for simulations, true random numbers matter because you don't want to skew things with any possible patterns.

True. We were however talking about cryptography and using /dev/random for generating a good key for wpa. Predictability is everything there. That said, you still don't want any weird paterns to show up, they are bad for preventing predictability.

> I used Linux as an example because I know how it works, not purely to push it. I guess I can see how you might have gotten that impression, though, so no hard feelings.

Ok, then my impression was wrong there, sorry. Oh, and your explanation is technically correct for all I know, but this is a case of 'good enough', not of 'scientifically correct' randomness.

At any rate.. the point of my original post was that not all Unix like systems have good sources for random numbers, others have good to excelent support for it. The later include systems like Linux, OpenBSD, NetBSD and FreeBSD.

Better colours

[Anonymous Coward]

http://shit.slashdot.org/article.pl?sid=04/11/05/2 143226

Re:Better colours

[Anonymous Coward]

One of these needs to be posted every article

Re:Better colours

[Anonymous Coward]

"Better colours: http://shit.slashdot.org/article.pl?sid=04/11/05/2 143226"

Ah, but is there a URL which doesn't give you either retarded Flash adverts or (depending on your tolerance for crap software) a popup message demanding you install stuff to read slashdot?

Re:Better colours

[SillyNickName4me]

> Ah, but is there a URL which doesn't give you either retarded Flash adverts or (depending on your tolerance for crap software) a popup message demanding you install stuff to read slashdot?

Yeah. Ever since the majority of ads here is flash based, the ad servers Slashdot uses made it onto my blacklist.

It is funny really. I did tolelrate the ads here for quite soem time sunce they were for a long time non-intrusive and somewhat targetted. Stick in flash and I just block the adserver (and yeah, I do have a flash player installed, and click to play also, but it is still annoying as hell)

So hello Slashpeople, if you read this, GET RID OF FLASH ADS!!!!! When you do and post about it, I'll unlist the ad servers you use.

Odds of implementation?

[IamGarageGuy 2]

The odds of Joe sixpack going the extra step of making a 20 character key is not good. WiFi setups are all the rage and now can all be broken into even after you spend an hour telling someone that they have to use WEP.

Re:Odds of implementation?

[EnronHaliburton2004]

WEP

Er, you mean WPA?

Re:Odds of implementation?

[IamGarageGuy 2]

doh! - temporal acronym overload

Re:Odds of implementation?

[fisgreen]

The odds of Joe sixpack going the extra step of making a 20 character key is not good. WiFi setups are all the rage and now can all be broken into even after you spend an hour telling someone that they have to use WEP.

Sadly, who needs to break into anything when so many leave their front doors wide open? I just moved into a new appartment complex. While waiting for my cable to get turned on, I thought I'd scan for networks, just for the hell of it. F'ing amazing: five APs detected, one WEP (not WAP) secured, four open. Of the open ones, three hadn't even changed the defaults.

Re:Odds of implementation?

[cbiltcliffe]

So:

1. Put up an ad in the mailroom for computer and network service in apartment number
2. A week later, start enabling WEP on the open routers.
3. Residents go ?????.
4. PROFIT!!

(Who knew this /. staple could ever be used in a sensible way?)

So it's just a bruteforce/dictionary tool...

[zaffir]

What's the big deal? Kismac has had this feature for a while. I hope i'm missing something.

Re:So it's just a bruteforce/dictionary tool...

[Tony Hoyle]

Kismet had a WEP sniffer - that has issues that allow you to derive the key from the data.

WPA (with AES, preferably) is a lot harder as it's designed so you can't go that way around.. you have to bruteforce.

A 128 bit key is only 16 characters (you *do* use non-ASCII in your keys I assume?). I usually try to go to at least double that.

Re:So it's just a bruteforce/dictionary tool...

[zaffir]

Notice i said Kismac, not Kismet. This new tool doesn't do anything special when attacking WPA. It isn't even the first to do this non-special thing.

Re:So it's just a bruteforce/dictionary tool...

[jrockway]

I like your sig :)

Re:What Morons

[Sarhosh Amiral]

It does not have to be cracked, MAC filtering does not prevent from others listening the network.

Re:What Morons

[Anonymous Coward]

Um, do you know how easy it is to spoof MAC addresses? Very easy.

I'll say it again - VPN

[Anonymous Coward]

Now that D-link and others sell routers with wireless and VPNs all in one box, just VPN with IPSec to your own network and that way you have everything running as securely as you can with a normal VPN. Sure, it's easier to sniff the traffic in the air, but it can still be sniffed on a wire too.

I'm all for this.

[Anonymous Coward]

Leaving my WAP wide open all the time allows experienced crackers to access all the best pr0n sites with ease via my connection. All I then have to do is check the logs and Voila! There they are! Saves me looking for them and having to wade thru the pop-ups and bogus sites!

Re:I'm all for this.

[Alby]

Plus it gives you plausible deniability when the RIAA/MPAA/DCMA/ETC come calling...

Re:What Morons

[StarWreck]

Before somebody flames me... get out your Laptop with a wireless card and logon to your Wireless Access Point. Go to the MAC-Filtering page, type in your exact MAC address for your laptop's wireless card. Now enable MAC-Filtering. Notice how you can still access wirelessly. Now change a single character in the MAC address you just typed out. You will never be able to access the internet wirelessly again. Unless you hard-wire yourself to it and change the address back.

First person to regain internet access after doing this without hard-wiring back into their LAN wins a monkey.

no good excuse

[Misanthropy]

there's not really any good excuse for a weak wpa key. My router will generate a random 128bit key.
Kind of funny. I have our wireless router locked down with a 128bit key and only accepting connections from mine and my roommates' MAC addresses. But one of my neighbors has a wide open access point that I can connect to whenever I wan't.
I don't really want to, but I could.

No real point to this post except that you should attempt even minimal security (Unlike my neighbor).

Re:no good excuse

[Anonymous Coward]

You should never access a neighbor's access point... ...except at night, to download porn, till dawn. Oh and with their permission.

"Pardon me, mind if I use your wireless connection so I can download porn and masturbate all night long?"

Re:no good excuse

[Anonymous Coward]

That actually made me laugh. good job AC

In addition to a cracker

[slashdot.org]

I would have liked to see a tool that will verify if your chosen key is 'secure' or not.

Would have made the crack software look a little less black-hat, to the uninitiated.

Just an idea.

Re:In addition to a cracker

[slashdot.org]

Just to clarify, before I'm getting wise-ass responses; I realize the crack software may be the ideal tool to confirm your key is A-OK. BUT, that software is not for the faint-of-heart.

Getting raw packets etc, is not something everybody knows how to do, but it would be great if they could verify their key was fine.

(and then that software could send the key back to the original web-site, so they can keep a list of fine keys that are now known, so they are not fine anymore. Yeah yeah, I know... ;-D)

Re:In addition to a cracker

[Harker]

I would have liked to see a tool that will verify if your chosen key is 'secure' or not.

But... Windows said it was secure. Isn't that good enough?? :P

Re:What Morons

[PedanticSpellingTrol]

Jesus christ, I hope you don't have a job in security. If all your packets are unencrypted, anybody can sniff them, see what MAC addresses are recieving traffic, and thus are on the whitelist. From there, it's a simple matter to spoof the MAC in software. This feature is built into linux, windows and OS X. The myth that MAC addresses are a universally unique identifier is dangerous and has to be dispelled.

This is why

[zakezuke]

This is why I setup a stand alone wifi network that when ever war-drivers discover my "wireless network" everything they visit gets redirected to goatse. The result, I've observed is usually a loud exclamation followed by the sound of screeching tires and burnt rubber.

Next i'll observe when I secretly host a wifi network near starbucks and replace everything with a small mirror of www.khaaan.com.

Re:This is why

[sketerpot]

If you want to get really evil, I assure you that some twisted people are perfectly capable of dreaming up even scarier things than goatse. For starters, do not click this link if you value your sanity. Does the phrase "penis bisection" pique your interest? If so, then remember what happens to people who go rashly clicking on links explicitly described as evil. *shudder*

Re:This is why

[zakezuke]

If you want to get really evil, I assure you that some twisted people are perfectly capable of dreaming up even scarier things than goatse

I don't know, hearing 20 laptops or so yelling "Khaaan! Khaaan!" I think is scarier than a penis bisection.

Re:This is why

[Anonymous Coward]

A mod like that is beautiful. The only cooler mods are a penectomy, nipplectomy, bilateral orchiectomy and clitorectomy. Do you have the guts to get nullified?

Re:This is why

[Anonymous Coward]

A mod like that is beautiful. The only cooler mods are a penectomy, nipplectomy, bilateral orchiectomy and clitorectomy. Do you have the guts to get nullified?

No sir, I don't :P

Re:This is why

[Anonymous Coward]

Nah. Once you have a man with two penises, you'll never go back.

Now, a pictures of a rectal prolapse, that's evil!

Re:What Morons

[RajivSLK]

Umm simple. I would just change the MAC address on my laptop to match the address in the WAP and reconnect. Many cards allow you to change the MAC address in software.

Now, where do I pickup my monkey?

Re:What Morons

[Anonymous Coward]

Depends on what activities you're doing while using the access point. If you're using VPN to your corporate network, it's already encrypted. Likewise with SSL.

Full text in readable form here

[Anonymous Coward]

I'm sorry, but here we read from left to right; not top to bottom.

Weakness in Passphrase Choice in WPA Interface

By Robert Moskowitz
Senior Technical Director
ICSA Labs, a division of TruSecure Corp

Use of PSK as the key establishment method

WPA and 802.11i provide for a Pre-Shared Key (PSK) as an alternative to 802.1X based key establishment. A PSK is a 256 bit number or a passphrase 8 to 63 bytes long. Each station MAY have its own PSK, tied to its MAC address. To date, vendors are only providing for one PSK for an ESS, just as they do for WEP keying.

When a PSK is used instead of 802.1X, the PSK is the Pairwise Master Key (PMK) that is used to drive the 4-way handshake and the whole Pairwise Transient Key (PTK) keying hierarchy. There is a straightforward formula for converting a passphrase PSK to the 256-bit value needed for the PMK.

This paper will look into the risks of using a PSK and particularly the risk associated with a passphrase-based PSK.

How the PSK is used in WPA and 802.11i

The PSK provides an easily implemented alternative for the PMK as compared to using 802.1X to generate a PMK. A 256bit PSK is used directly as the PMK. When the PSK is a passphrase, the PMK is derived from the passphrase as follows:

PMK = PBKDF2(passphrase, ssid, ssidLength, 4096, 256)

Where the PBKDF2 method is from PKCS #5 v2.0: Password-based Cryptography Standard. This means that the concatenated string of the passphrase, SSID, and the SSIDlength is hashed 4096 times to generate a value of 256 bits. The lengths of the passphrase and the SSID have little impact on the speed of this operation.

The PTK is a keyed-HMAC function using the PMK on the two MAC addresses and the two nonces from the first two packets of the 4-Way Handshake. This is why the whole keying hierarchy falls into the hands of anyone possessing the PSK, as all the other information is knowable.

The Intra-PSK attack

The normal practice is to have a single PSK within an ESS. To generate any PTK, a device only needs to learn the two MAC addresses and nonces (and the selected ciphersuite). All of this is available in the initial exchange, from the ASSOCIATE through the 4-Way Handshake. Any device can passively listen for these frames and then generate the PTK. If the device missed these frames, it can send a DISASSOCIATE against the STA and force the STA to perform the ASSOCIATE through the 4-Way Handshake again.

Thus even though each unicast pairing in the ESS has unique keys (PTK) there is nothing private about these keys to any other device in the ESS.

The offline PSK dictionary attack

A station that does not know a passphrase-based PSK can attack it with an offline attack. This is effective for an outsider where there is a single PSK in the ESS, or an insider where there are unique PSKs.

The 802.11i standard points out that:

A passphrase typically has about 2.5 bits of security per character, so the passphrase of n bytes equates to a key with about 2.5n + 12 bits of security. Hence, it provides a relatively low level of security, with keys generated from short passwords subject to dictionary attack. Use of the key hash is recommended only where it is impractical to make use of a stronger form of user authentication. A key generated from a passphrase of less than about 20 characters is unlikely to deter attacks.

The PTK is used in the 4-Way handshake to produce a hash of the frames. There is a long history of offline dictionary attacks against hashes. Any of these programs can be altered to use the information in the 4-Way Handshake as input to perform the offline attack. Just about any 8-character string a user may select will be in the dictionary. As the standard states, passphrases longer than 20 characters are needed to start deterring attacks. This is considerably longer than most people will be willing to use.

This offline attack should be easier to execute than the WEP attacks.

Using Ran

Re:What Morons

[StarWreck]

The deal was for you to actually do it, not explain how to do it. /monkey bitch slaps you

Re:What Morons

[chizu]

"ifconfig wlan0 hw ether [mac address]" sets your wlan card's mac address under Linux. There is probably a way to do so under Windows as well.

Re:What Morons

[iamnotacrook]

you need to brute-force check each MAC adress. there are ways to make this harder in the router.

don't blame WPA

[nbert]

...if your key is asdf - the attack is based on a dictionary. This weakness relies on human nature after all.

Btw: The Tips and Tricks section of this newsletter is a good ressource if you want to create passes which are harder to guess.

Re:don't blame WPA

[nbert]

arghh - let's blame my caffeine consumption...

Here's the a correct link

Re:don't blame WPA

[PMJ2kx]

...if your key is asdf - the attack is based on a dictionary. This weakness relies on human nature after all. So, write a generator to make a random key so that it's not just ASDF...just make sure to memorize it...ya know, just in case... ...or, if you're like my not-so-bright friend, write it on a stickynote and leave it in a book labeled "passwords & stuff"...

Re:don't blame WPA

[Anonymous Coward]

dd if=/dev/random bs=1024k count=1 | md5sum

there is a 16 byte (represented with 32 hex chracters) that is probably farily random. If you need ascii characters, hmm, play with this:

echo `dd if=/dev/urandom bs=1024k count=1 2>/dev/null | strings -1 -es` | sed s/\ //g | dd bs=1 skip=$RANDOM count=8 2>/dev/null ; echo

I couldn't figure how to drop the new line stuff, so that's why the echo and the sed(assumes a bash shell). It also makes some asumptions, like it will have at least 32768 characters ($RANDOM says it give a number between 0-32768). Change that 8 characters to something larger if you want more characters(20 or 60 or what ever). in the strings, change the "-es" to "-eS" to allow 8 bit characters, as is right now it will just do the 7-bit characters. (although on my system it didn't seem to play well with that). Change the /dev/urandom -> /dev/random to not use the psuedorandom.

I will say running from urandom I have gotten files from the current dir listed as well as some of the environment variables, so beware.

Re:don't blame WPA

[/dev/trash]

If an attcker has physical access he's already won. Having your wireless password written down isn'y gonna make it easier for some kid looking for a free ride.

Re:don't blame WPA

[nbert]

I like the semi-bright approach:

My passes (coming from /dev/urandom) are stored in my encrypted home dir on my box. So if I need one of them 'on the road' I just ssh in and look it up (having putty on a stick helps a lot if you have to deal with Windows btw).

IMO that's a convenient tradeoff between remembering them all (impossible) and having them on notes (insecure).

Re:don't blame WPA

[mattyrobinson69]

i use wep 128bit, its a right bitch because it uses hex. using hex means i cant whack the keyboard randomly until ive got my 128bits. i just had to type in numbers randomly and occasionally press one of abcdef. That and mac address filtering kept out that bastard kevin leeching (didn't turn on any security at first because i forgot wireless was enabled).

Re:What Morons

[Anonymous Coward]

the article talks about first capturing packets with Ethereal... don't you need to be on-net to do that? so how can you first be on-net if MAC filtering is on? you'd need to know a valid MAC ahead of time... pretty unlikely. am I missing something?

Re:What Morons

[wcdw]

NOT really a good idea to start a thread about morons, and then act like one.

_YOUR_ wlan card may have the MAC address burned into it. Once ALL NIC did. I think it was more than 10 years ago that I saw my first NIC that DID NOT HAVE a MAC address (it was all zeroes, and expected to be set in software).

_MY_ wlan card will _CERTAINLY_ let me change the MAC address - under Linux _or_ Windows.

http://www.theboyz.biz/Computers, parts, electronics, small appliances and more!

Ho hum

[Realistic_Dragon]

Guess it's not time to abandon treating all wireless hosts as bastions and using SSH to tunnel/authenticate just yet then.

Treat wireless just like you do a student network and everything will be fine.

Re:Ho hum

[Dr. Evil]

Note that WPA is just like WEP but with quickly rotating keys and more secure key exchange. Yeah, you can't crack it in real-time to get on the network... but if you listen to the vendors carefully, they'll even say it... "Authentication, Authorization.... " But never will they formally say "Secure encryption of data"

You can decode everything but the key exchange off-line.

VPN software is the only way to go. The wireless vendors are liars.

Does anyone want to comment on WPA2? Does it require new hardware?

Re:What Morons

[wcdw]

Once again I draw your attention to the problems with making conclusions about the entire world based on your one example.

This is a *trivial* exercise, and certainly does not require any hard-wiring.

http://www.theboyz.biz/Computers, parts, electronics, small appliances and more!

Re:WPA Keys

[Olmy's Jart]

Yes... Several..

Do your homework. Look up Supplicant, XSupplication, HostAP, 802.11i for Linux, 802.1x for Linux, etc, etc, etc... Lots of things going on.

ITMT... This crack is only for weak keys with WPA-PSK. Not applicable to WPA enterprise or WPA2.

Re:What Morons

[EnronHaliburton2004]

My Linksys card allows me to change the MAC on Windows... it's trivial to do.

The other poster sowed how easy it is to do this in Linux.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:What Morons

[Anonymous Coward]

i love when idiots like you post on nerd sites and make an ass of yourself.. you should have posted anonymously, your nerd creds have been lost, you can never show your face here again as StarWreck.. time to make a new username or never come back, you ruined it

and while yes this is a troll.. its not a pure troll.. had you posted only your first post then replied to the replies with something like "oh i wasn't aware of that, sorry, i guess i was wrong"... then you'd be fine.. but you keep replying saying you are right and everyone else is wrong.. when everyone else is right and you're wrong..you're probably not stupid, you made a simple mistake, but then you acted like an ass about it and now you ruined your slashdot name

Re:What Morons

[arth1]

you need to brute-force check each MAC adress. there are ways to make this harder in the router.

No, you don't have to do this. Once the WEP key is broken (or if there is no WEP key, just MAC filtering), you simply listen to the traffic to get a MAC address that's allowed, and use that.

Regards,
--
*Art

Re:What Morons

[Anonymous Coward]

but your mac address is still NOT encrypted

D-Link Software generates 60-digit pass keys.

[Futurepower(R)]

D-Link's install software for the AirPlusXtremeG WiFi adapters generates a 60 digit random hexadecimal number for use as a pre-shared key.

The Daily Mirror is one of the United Kingdom's largest newspapers. Here is their front page on the day after the election (PDF file): Daily Mirror Front Page: How could 59,054,087 people be so dumb?.

Re:D-Link Software generates 60-digit pass keys.

[Glendale2x]

The Daily Mirror is one of the United Kingdom's largest newspapers. Here is their front page on the day after the election

Could someoneexplain to me what that has to do with WPA? or D-Link keys?

Re:D-Link Software generates 60-digit pass keys.

[pesc]

Nevada's Ministry of Propaganda for the Democrats: UNR

Could someone explain to me what that has to do with WPA? or D-Link keys?

Re:D-Link Software generates 60-digit pass keys.

[Glendale2x]

Do you know what a sig (AKA "signature") is? I want to make fun of my school in my sign, and many sigs are offtopic, since you want to make that point.

You are just ignorant.

Relax dude!

[Anonymous Coward]

Dude, I would seriously lay off the 18 cup a day coffee diet.

Suggestion

[cuteseal]

From reading all the threads and flame wars going on here, it appears that WEP, WPA and even MAC address filtering is easy to crack, if someone was determined enough to do it.

So, I know it's not foolproof, but does anyone have suggestions on how to increase wireless security?

1. Regularly change WEP keys?
2. Use a proxy server to access internet, and disable direct access via access point?
3. Turn off router and computers when you're not using them?

Any others?

Re:Suggestion

[Anonymous Coward]

IPSec and SSH tunneling.

Re:Suggestion

[StarWreck]

The lesson learned here is to use multiple layer security. As I said in an earlier post, 1-layer security is always a bad idea. Use a good length WPA (or WEP) as well as Mac-Filtering. Encryption and Filtering both have their weaknesses but combined they are fairly effective.

Re:Suggestion

[slashname3]

The best thing you can do in addition to using WEP, changing keys, and locking down the MAC addresses allowed, is to use ssh or VPN software to encrypt your connections. If someone spends enough time to crack WEP and spoof a MAC address then the most they can get is access through your access point. They would have to break ssh or VPN to look at your data. Of course you would need to have tools in place to identify a man in the middle attack to prevent them from spoofing your connections.

Of course if someone spends that much effort just to break into your wireless network you either have something really important or they are have way to much time on their hands. (and I doubt if anyone has anything that important on their network....)

Re:Suggestion

[igrp]

From reading all the threads and flame wars going on here, it appears that WEP, WPA and even MAC address filtering is easy to crack, if someone was determined enough to do it.

Well, there are different schools of thought when it comes to SoHo/low bandwidth WAN access security.

You are attempting to lock your network down so that a potential attacker cannot use your connection. The other approach lock your network down just enough to make a cracker not want to bother and to move on to the next, easier target (ie. your neighbors' access points).

The former approach generally works just fine if your goal is to deny a potential attacker access to your network bandwidth. It won't really stop a determined attacker who isn't just in it for a free-ride but who wants to steal specific data. If that's part of your threat model, chances are wireless isn't really for you. The downside is that this is pretty inconvenient. And since convenience is the big selling point when it comes to wireless networking, most people just won't take that route.

Those people who have WEP and MAC address filtering enabled, basically want to protect themselves against random, unsophisticated wardriving. It won't help defend against a determined attacker and probably won't even scare off the teenager next door with too much time on his hands. The point isn't really to have good access security. It's just to raise the bar enough to be unatractive enough of a target. Think of it as a "I don't have to outrun the bear, I just have to outrun you" scenario.

Re:Suggestion

[pebs]

1. Regularly change WEP keys?
2. Use a proxy server to access internet, and disable direct access via access point?
3. Turn off router and computers when you're not using them?

All good ideas.. A bit of extra work, but worth it if you really need to be secure. It's also important that someone on your WLAN can't access anything on your regular LAN, so be sure to setup a firewall that prevents this. Probably what I would do is setup either a web proxy over SSH or a VPN. But someone on the WLAN can only access the SSH port or the VPN port. Also, throw in some MAC address checking for extra measure.

Everything can be exploited if someone is determined enough, but the point is to make it as difficult as possible.

Re:Suggestion

[eggboard]

WPA isn't easy to crack when you choose a good key. Just invent a passphrase like "My d0g!! has f6666%%&%__seas" and you're safe as houses.

Re:Suggestion

[Homology]

So, I know it's not foolproof, but does anyone have suggestions on how to increase wireless security?

My home firewall is an OpenBSD box that is my access point as well. I use IPSec to setup VPN to secure my wireless network. Only authenticated IPSec traffic is permitted, so all a war driver can do is to DoS my wireless network.

If setting up IPSec is too much work, one can use OpenVPN that has a Windows client as well.

If you just want to prevent unauthorized usage of your wireless network, you can authenticate using authpf.

All the soloutions above assumes that you uses OpenBSD as an access point. OpenBSD can now support Atheros wireless chipset (for 802.11a), and soon 802.11g will be supported as well : Atheros HAL layer.

Just name all your specific MAC addresses

[NotQuiteReal]

This will also help secure your network.

How many home networks really need to allow random MAC addresses access?

Re:Just name all your specific MAC addresses

[hsidhu]

ummmm how hard is it to sniff the traffic, and get the MAC addess that is allowed and then spoof it?

Re:Just name all your specific MAC addresses

[zakezuke]

How many home networks really need to allow random MAC addresses access?

How many home users know what a MAC address is?

Re:Just name all your specific MAC addresses

[Anonymous Coward]

Good thought, but no good for WPA and in general.

For WPA: Part of the encryption process apparently obfuscates the MAC, hence access points only being able to do AP-to-AP bridging using WEP or no encryption at all.

In general: it's way trivial to sniff a valid MAC and then spoof yours.

Re:Just name all your specific MAC addresses

[IHateSlashDot]

You're kidding right? MAC filtering provides absolutely no added security. Once the encryption is broken, spoofing a MAC address is trivial.

Re:Just name all your specific MAC addresses

[Anonymous Coward]

How many home users know what a MAC address is?

Let me guess...

A: The street address of an Apple Macintosh?
B: The serial number of an Apple Macintosh?
C: Address of the closest MacDonald's?
D: Something that can be changed under Network Device Settings?

Re:Just name all your specific MAC addresses

[Anonymous Coward]

E: Mcintosh inc makers of fine audiophile equipment.

Re:Just name all your specific MAC addresses

[Anonymous Coward]

I'm going to have to go with B, because C would be a BigMAC address...

Re:Just name all your specific MAC addresses

[DieByWire]

How many home users know what a MAC address is?

I've got three MACs at one address. What am I supposed to do, make up an apartment number for each one?

Re:Just name all your specific MAC addresses

[Anonymous Coward]

ummmm how hard is it to sniff the traffic, and get the MAC addess that is allowed and then spoof it?

Trivial.

Re:Just name all your specific MAC addresses

[Moloch666]

Actually you can get the MAC address before cracking the encryption. That information is not encrypted.

Re:What Morons

[kg4gyt]

Sure you can change your MAC Address, but then you have to change the MAC Address to one that is valid. If no one is on the network, but you can hear it by simply wardriving, its not going to do anybody any good.

Re:What Morons

[Anonymous Coward]

"the article talks about first capturing packets with Ethereal... don't you need to be on-net to do that?"

No, you can set your card to 'monitor' mode, and pull other packets out of the air. MAC address can be had in these...you just change your card's MAC address, and poof!, there goes MAC address filtering!

record now crack later

[pronobozo]

How about technologies of the future that you can just wait around for. I am sure in 5 years the hardware then will be able to crack stuff now in a matter of seconds. So why not record now... be patient(5 year wait).. and then crack.

Re:record now crack later

[shuut]

Depending on your browser some old browser might only support encryption with less than 128 bit keys in SSL, looking at the specs that's basically DES and RC2, so if you sniffed an encryption message sent through by one of these less than 128bit browser you can break it in matter of days.

Re:What Morons

[Anonymous Coward]

_did_ it _really_

Asside: WEP = Wired Equivalency Protocol

[KillerCow]

As an aside to the above point, the original "WEP" stood for "Wired Equivalency Protocol." They chose that because it acknowledged that wires weren't inherently secure either. It's name didn't claim security at all... just that it was equivalent to a wire. The inside joke was that that didn't mean anything from a security standpoint either.

Correction: 64 4-bit hexadecimal keys

[Futurepower(R)]

Correction: 64 4-bit hexadecimal keys, for 256 bits total. According to the article, not breakable.

What is Slashdot coming too?

[Anonymous Coward]

I know traffic has been declining to this site but please have a little dignity left. Posting cracks on slashdot? What next, hosting the latest music, movies and software. I would hope the moderators would do a better job sifting through stories. Lots of good stories are getting rejected while dupes and stuff like this gets posted all the time. It's just a shame to see this site suffering from the same problems big media conglomerates have.

Re:What is Slashdot coming too?

[Anonymous Coward]

So you think there should be no news about security ever? So that everyone would feel safe? Vulnerabilities should never be mde public when discovered? You think sysadmins should no know that their WEP/WPA wireless network is not secure?

Not again "security through obscurity", please...

One more question though... do you work for MS?

Phew i'm safe!

[Anonymous Coward]

Mine is qwerty

Re:Phew i'm safe!

[Anonymous Coward]

Mine is azerty, so hah.

Takes a load off my mind, in a way.

[Anonymous Coward]

I had started on a dictionary cracker for WPA keys, but it was for a class project and once the class was over, the project pretty much was as well. Not being much experiened in POSIX I/O for network and wireless interfaces, I had no idea what to do to put in those parts and was kind of torn on whether to take time out to learn it. It's still on sourceforge if anyone wants to finish it, i.e. add in the bits for acquiring actual packets.

Maybe we can modify it for AES/CCMP keys. They still use passphrases, right?

Wired security vs. wireless security

[thedillybar]

Just because you have a wired connection, doesn't mean you can assume no one is snooping on the wire.

Anything confidential needs to be encrypted with VPN, SSL, or something similar. Period.

Sure you can turn WEP or WAP on...but don't stake all your data on it. Use what's tested and trusted by the rest.

Re:Wired security vs. wireless security

[Moloch666]

Scary thought. I look around my little room and the LAN cable running along the wall. Wonder... "Could someone be tapping into my connection right under my nose?" If they were in my house, why not just take the computers.

I'm just messing with you, I know you meant larger enviroments.

Re:What Morons

[hkb]

_LETS_ continue to _TYPE_ like this _A LOT_. it's _PRETTY_ _COOL_ and gets my _POINT_ across really _WELL_!!!!

Link to TinyPEAP's WPA Cracker utility

[OneNonly]

For those who are interested in checking out your own security (I'm just about to do this!) the WPA Cracker that has been released is available here: WPA Cracker.

I wonder if 13 characters is short :S

Re:Link to TinyPEAP's WPA Cracker utility

[Anonymous Coward]

I wonder if 13 characters is short :S

I notice that your slashdot name and number total 13 characters......

What about unsecured networks?

[porkUpine]

Until people start securing their wireless networks with SOMETHING, wireless will always have a bad reputation. As nice as it would be, we aren't allowed to use wireless in office... period. BTW, I'm surfing /. from my neighbors unsecured WAP. *Sigh*

Ok, MAC spoofing seems "trivial"...

[NotQuiteReal]

...but I live out in the burbs.

Point taken - I won't put anything important on my wireless access, but then again, I am about as likely to have someone out here care about spoofing me as I am to have some gang-bangers drive the 50 miles to my "hood" and rough me up.

Computers are becoming more like regular life - assume someone will someday see what is on your computer, just like your are probably being watched by security/traffic cameras all the time.

heh, "play nice", even when you think no one is watching and you will be ok.

Re:Ok, MAC spoofing seems "trivial"...

[Anonymous Coward]

What did you have for dinner? What is the color of your socks right now? What are you thinking right now?

I expect detailed reports from now on, every day. And don't dare to lie! You DO play nice, do you? Huh? HUH?

Re:Ok, MAC spoofing seems "trivial"...

[NotQuiteReal]

Ok.

I haven't had dinner yet, but I will probably just have fast food - since I have been painting (non-artistic) all day and noone feels like cooking.

I always were white socks, like all geeks, and am therefore puzzled about the black stuff I pick from under my toenails, since... well duh.

I am not thinking much right now, since I've knocked back a few - perfect for making slashdot toasts, erm, I mean posts.

Basically "playing nice" means being honest - if you like porn, by all means have it on your computer, but if you think it is "dirty" and do it anyhow, shame on you for being a hypocrite.

If you "play nice" - i.e. are honest with yourself and others, you will never be embarassed by what is on your computer - you might be pissed that someone violated your "privacy", but it won't ruin your political career or bring down your TV evangelist empire or anything like that.

That is all I meant - if you were to breach my computer you might find some financial info that I'd rather not be public, but other than that, I'd not be embarrassed (unless you critiqued some of my coding) by what you found.

What are you smoking?

[Anonymous Coward]

I am a part-time sysadmin for a small company. My most important duties are things other than administration. Yet, all the administration in the company is done by me.

This was an EXTREMELY important piece of information to me. I had been under the impression that "anything WPA has not been broken yet, and is inherently more secure than WEP".

Now I need to figure out how to reconfig those APs to talk 802.1x to a server, which is going to be so not fun, but a lot more fun than having discovered a fait accompli break-in.

Stop the Press !

[Anonymous Coward]

Newsflash - Weak Keys Crackable !

uh, no shit...

Re:What Morons

[Anonymous Coward]

_YES_ it did _REALLY_

Re:What Morons

[wfberg]

MAC adresses are universally unique identifiers, except for a few duplicate runs in cheap-ass brand NICs.

It's just that they cannot be authenticated in any way. It's like allowing only people who claim to be you on your network, rather than people who can prove it in some way.

What are "short" WPA keys supposed to be?

[rpp3po]

It is easy to see, that the original poster of this story has no clue about encryption. There are several misconceptions in his posting:

1. He writes: "WPA is the replacement for weak WEP keys in the original 802.11b specification". This is wrong. "weak key" ist a crypographic term for - wonder - weak keys, like 128 bit, consisting of 1's only (1111111111111...). For like 30 years, even WEP, has taken measures to prevent this kind of keys during use. WEP's problem in fact is the deterministic generation of IV's of the keystream, not weak keys.
2. "Moskowitz advised choosing passphrases of more than 20 characters or generating random keys of at least 96 bits, but preferably 128 bits." That's also misunderstood. The PSK (pre shared key) even when not using 802.1X is always 256-bit. It's generated -from- a passphrase that you type in. A passphrase like "abc" e.g. contains less than 16 bits of security. So a WPA key generated from the passphrase "abc", although still being 256-bit, can be cracked within the time of a 16 bit brute force attack. This is done by simply generating WPA keys from all passphrases between "aaa" and "zzz". So you always use 256 bit keys (PSK's), but they can be generated from much smaller passphrases.
3. "each user gets a long WPA key". See above. The keys are always the same size of 256 bit. When using 802.1X there is only maximum "randomness". That's the difference. It think the poster still thinks that WPA works like WEP where you actually use different key lengths.

One could think that I'm very picky about his words. I think not. Especially in cryptography it is important to know exactly what part of a cryptographic chain you're talking about, when talking about weaknesses. TinyPEAP seems to be just a tool for people like the original poster and script kiddies, who are in fact NOT knowing what they are talking about. It's just a bruteforce tool to try out WPA passphrases. This is supposingly faster for people using short passphrases than bruteforcing keys directly.

Re:What are "short" WPA keys supposed to be?

[Anonymous Coward]

WEP has several problems. Deterministic IV generation is not one of them. To be precise, sequential IVs are preferable to random IVs because you can effectively avoid using the same IV twice. The IV is transmitted in the clear, so you don't need knowledge about IV generation to get the IV.

WEPs main problem is that the space from which IVs can be chosen is much too small. That, combined with a user supplied key which is directly used for encryption instead of just securing the exchange of random keys, means that you can't avoid reusing the same key. The RC4 algorithm used by WEP becomes vulnerable when the same key is used twice.

Due to the nonexistent defense against replay, attackers can create arbitrary amounts of traffic on encrypted networks. That means they can provoke IV reuse. Sequential IVs can be used to reject frames which are encrypted with IVs that the recipient has seen before, thereby foiling replay attacks.

Re:What are "short" WPA keys supposed to be?

[eggboard]

I'm the original poster. You're reading what I wrote with the lack of knowledge necessary to comment on it.

1. All WEP keys are susceptible to nearly the same degree of being broken by collecting enough data passively. Thus, they are all weak. From a definition of weak keys at an online dictionary: "In the extreme, a poor cipher design is simply one with a very large number of weak keys."

2. No, you're misreading this, too. Moskowitz (see his paper) is talking about the seed data, not the resulting way in which it's represented. The lack of randomness in seed data is the problem. So if you take 16 bits of data and turn them into a hex WPA key, it doesn't matter whether it's represented as 256 bits. The whole problem is the algorithm by which it's processed. You need to start with at least 128 bits of data (into hex) that are non-dictionary, non-weak. (In this sense, weak is much more limited.)

3. Sigh. Each user gets a key that has a full 256 bits of randomness.

You are being picky about your words incorrectly.

Re:What are "short" WPA keys supposed to be?

[rpp3po]

1. Wrong. "Weak key" is a well defined cryptographic term. See e.g. http://en.wikipedia.org/wiki/Weak_key . What you're describing is just a bad cipher. If all keys are bad to the same degree the keys aren't called weak but the cipher.
It would be different if there were stronger keys additionally to the huge number of weak keys in the same set of the cipher you are talking about. This is not the case.
2. The seed's randomness is exactly what I was talking about. It's not what you are saying, that the "whole problem ist the algorithm". There is no algorithm in the world which can extract more than 16 bit randomness from a 16 bit seed (passphrase). So that's NOT the algorithm's problem, but people using too short passphrases.
What do you mean by "128 of data (into hex)". It doesn't need to be hex. 128 bit randomness are 4 bit per digit in hex. Using average passphrases you get about 2.5 bits per character, but more if you use the whole character set. So using passphrases or hex numbers gets you the same thing (in bits).

Re:What are "short" WPA keys supposed to be?

[Anonymous Coward]

Implementations of WEP which avoid weak keys are known as WEP+. Early WEP implementations did use weak keys (in the cryptographic sense). The small percentage of weak keys is the reason why you have to capture a lot of traffic to find the key.

Re:What Morons

[golgotha007]

I was under the impression that all NIC's had a hard coded MAC address.

You can change the MAC address via software in the NIC driver, but that doesn't physically change the hardware MAC, it simply changes what the driver reports the MAC address as being.

Am I wrong here?

Re:What Morons

[zzyrc]

The frame control that contains the MAC header in an 802.11 packet is always unencrypted. So the list of MAC addresses is available at once, before key cracking.

Re:What Morons

[Anonymous Coward]

* MAC adresses are universally unique identifiers, except for a few duplicate runs in cheap-ass brand NICs.*

actually.. they're not. it's not guaranteed in any way(and would a manufacturer run out of them he would just start from the beginning again).

Re:What Morons

[Anonymous Coward]

They're supposed to be unique. If you have two cards with the same MAC address on the same ethernet, the network will not work as expected. The manufacturer prefix is a way of delegating responsibility, not a way of saying "Your ID is unique, what you do beyond that is your own damn business."

Re:What Morons

[Olmy's Jart]

You are correct. You are wrong. Even if some cards have immutable "burned in MAC addresses" (I not aware of any) the fact remains that most of them allow you to set the MAC address. The bad guy merely has to buy the card that lets him do what he wants to do. Even if you only buy fixed address cards, he's not so restricted. So, even if only one model of one brand allowed this technique, you would still be screwed. And, AFAIK, it's the majority of cards which allow it, not even a minority. And, yes, it really does change the address on the air.

Re:What Morons

[PalmerEldritch42]

I'm not aware of any NICs that are totally un-alterable MAC-wise. There may be some out there, but none of the consumer level ones I've used have any trouble changing MAC addresses. As one previous poster stated, in Linux, you can use this command:

ifconfig wlan0 hw ether [mac address]

In Windows, there is a nice program called Mac Makeup to do it for you.

Both of these methods do work for wireless cards. I tested it fairly extensively when I setup my own wifi network.

Re:Aside: WEP = Wiretap Equivalent Privacy

[Anonymous Coward]

That's the inside joke.

WPA er Old News!

[fogez]

KisMAC has had this function for a long time. Someone used it at Hope 2004 to their wifi key. In addition, Josh Wright has had a working copy available for linux for some time. The LiveCD from Remote-exploit.org (Auditor) has included this tool for about a month now. This is not new...

Re:What Morons

[Anonymous Coward]

There are more NICs then there are MAC addresses. They are not unique identifiers. I have personally had do deal with MAC conflicts on Apple PowerMacs and 3com cards. The very fact that they are not guaranteed to be unique by the manufacturer should clue you in to their usefulness as a security measure.

Re:What Morons

[Anonymous Coward]

There are more than 2^48 NICs? That's 281,474,976,710,656 if you don't have a calculator handy.

Know Thine Adversary

[bill_mcgonigle]

You're kidding right? MAC filtering provides absolutely no added security. Once the encryption is broken, spoofing a MAC address is trivial.

You're assuming your adversary is a determined attacker, not your neighbor who's too lazy/clueless to choose his access point in his laptop's configuration utility. MAC filtering works perfectly well in this case.