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

[cross-posted from ba.internet]

On Sat, 05 Aug 2006 11:02:59 -0700, Jeff Liebermann
<jeffl@comix.santa-cruz.ca.us> wrote in
<eqm9d2d7shdbnnja33trmdb3k3mh4lv97v@4ax.com>:

>John Navas <spamfilter0@navasgroup.com> hath wroth:
>
>> "in the longer term WEP devices will be phased out"
>>
>>Optimistic and unrealistic (IMnsHO), even worse than Microsoft trying to
>>phase out Windows 98. I think WEP devices will be in significant usage
>>for many years to come, and that few low-end vendors will be brave
>>enough to switch to devices without WEP support in the foreseeable
>>future.
>
>Yep. WDS bridging does not work with WPA. The best that can be done
>is WEP. As long as WDS is around, so will WEP. It's not just WEP for
>the wireless routers, one has to use WEP on the clients in order to
>use WDS. Bummer.

According to
<http://expertanswercenter.techtarget.com/eac/knowledgebaseAnswer/0,295199,sid63_gci1104925,00.html>
current versions of the following products support WDS with WPA (albeit
likely only with products from the same vendor):
* Apple Airport 4.1
* Sveasoft Alchemy
* Belkin 802.11g Wireless Network Access Point
* 3COM OfficeConnect Wireless 108 Mbps 11g PoE Access Point
* Corinex Wireless to Powerline Router G

>Interestingly, WEP+ (Lucent/Prism) seems to do quite well against the
>attacks I've been trying. I'm still tinkering but I just can't get a
>sufficient number of "interesting" initialization vectors from a WEP+
>router to be useful for cracking the WEP key.

Although I've not yet heard of a successful attack against WEPplus
(WEP+), which works by avoiding "weak IVs", I still have objections:
* Proprietary
* Lack of peer review
* Only fully effective with WEPplus products at both ends, which isn't
easily enforced, a serious limitation.
* Doesn't protect against replay (and DoS) attacks.

--
Best regards,
John Navas