WDS with WPA, WEPplus (Was: SOHO LAN Question re: MAC addresses)

[cross-posted from ba.internet]

On Sun, 06 Aug 2006 02:52:50 -0700, Jeff Liebermann
<jeffl@comix.santa-cruz.ca.us> wrote in
<2edbd2po1nd6sv716csqrssggce3j19kfg@4ax.com>:

>For your entertainment from Tom's Hardware:
>
>http://www.tomsnetworking.com/2006/0...page4.html#569

"Thnx to Buffalo Technology for their help on this question."

I'm not terribly surprised -- Tom's has sadly fallen far from its
iconoclastic roots, and is much less useful than it used to be.

>Some notes on WEP+
>http://www.oreillynet.com/cs/user/view/cs_msg/26023

I think that does a terrible disservice -- just because WEPplus wasn't
in his particular test case doesn't mean that it is safe even now, much
less in the future.

<http://www.wi-fiplanet.com/columns/article.php/1443911>

Proprietary Versions

Proprietary WEP flavors have come to market, most notably, Agere's
WEPplus. According to the company, WEPplus remedies the
initialization vector problem. However, is the proprietary version
more secure than conventional 128-bit WEP?

"The word on the street is that WEPplus is truly more secure than
other WEP extensions," says Greeson. "Given the use of a key
generation algorithm that avoids weak keys, WEPplus is more difficult
to crack." Greeson warns that Agere claims its WEPplus only "reduces
the vulnerability," as opposed to eliminating it.

"Orinoco's WEPplus is clearly an interim innovation, which definitely
provides added security over the basic WEP protocol," says Sabharwal.
"It is capable of defeating AirSnort, though ultimately it implies
that both the client and access point devices have implemented
WEPplus. [AirSnort is a hacker tool that enables someone with a
PRISM2-based 802.11b card, Linux 2.4 kernel based system, and WLAN
drivers to guess a network's WEP key after passively gathering
between 100MB-1GB of data.] That is why in the longer run the
industry must move towards truly standardized security solutions."

<http://www.netstumbler.org/archive/index.php/t-12241.html>

devine 08-30-2004, 07:46 AM

The other point which I see to the whole WEP breaking argument is in
regards to the weak IV packets. Specifically, how newer firmware
tends not to allow weak packets to be transmitted. This then begs the
question about users and the firmware upgrade cycle. That of course,
just leads back to questions revolving around overall security and
how often audits and reviews are done. However, the bottom line is
that since fewer and fewer firmware releases allow weak packets to be
transmitted, even as the attacks become more efficient, there is
point of diminishing returns.

That's true. Indeed, before the KoreK attacks came out, tools like
AirSnort were mostly useless because almost all post-2002 wifi
hardware implements wepplus (weak IVs filtering).

Now, with these new attacks, one can crack a WEP key with 0 "weak
IVs" (or "interesting IVs", in the AirSnort terminology). And one
needs much less packets; 500k unique IVs are sometimes enough to
crack a 104-bit WEP key.

Of course, it's possible that hardware vendors will start providing
firmware updates to thwart the KoreK attacks; but as most are
upgrading to WPA/WPA2, this appears quite unlikely.

devine 08-30-2004, 09:47 AM

... It usually takes a couple of hours to gather enough packets, ...

There's simply no good reason not to use WPA.

>>Although I've not yet heard of a successful attack against WEPplus
>>(WEP+), which works by avoiding "weak IVs", I still have objections:
>>* Proprietary
>
>WEP Plus is in any access point with a Prism chipset since 2002,
>Intersil chips since firmware 1.56, Orinoco since 8.10. Nothing in
>Marvell, Atheros, or Broadcom.

I don't see a relevant point.

>>* Lack of peer review
>
>Lack of complaints?

Again, I don't see a relevant point.

>>* Only fully effective with WEPplus products at both ends, which isn't
>>easily enforced, a serious limitation.
>
>True. If you're going to attack the client, instead of the access
>point, that might be important. Normally, the infrastructure client
>will not respond to the usual ARP injection or auth/assoc pounding. An
>ad-hoc client might respond. Of course, one could passive sniff the
>client and try to accumulate the necessary IV's (the old way). If the
>AP has WEP+, methinks it's sufficient protection.

Methinks you're way too complacent. Complanency is antithetical to
security.(c) That you don't know of an easy method doesn't mean that
one doesn't exist.

>>* Doesn't protect against replay (and DoS) attacks.
>
>True. It doesn't help against a replay attack. However, there are
>far easier ways to do DoS with wireless (e.g. cordless phone).

The three laws of prediction:

1. When a distinguished but elderly scientist states that something
is possible, he is almost certainly right. When he states that something
is impossible, he is very probably wrong.

2. The only way of discovering the limits of the possible is to
venture a little way past them into the impossible.

3. Any sufficiently advanced technology is indistinguishable from
magic.

-Arthur C. Clarke

--
Best regards,
John Navas