[MLB-WIRELESS] DNS and Locfinder

Andrew Griffiths andrewg at d2.net.au
Thu Nov 28 17:52:48 EST 2002


Hi evilbunny,

evilbunny wrote:
> Hello Andrew,
> 
> AG> Route poisoing is easier to deal with, such as authenication between 
> AG> two/more peers (hmmm, I think replay attacks are a problem though, its 
> AG> been a while since I played with various protocols in respect to that), 
> AG> striking out route updates, (e.g a interior router saying its got routes 
> AG> for another area when you know it can't have.). *shrug* preferably 
> AG> though, imo we should be aiming to have a focus to making things more 
> AG> secure than currently what we've got with the internet.
> 
> Yes and no, routing protocols assume a secure link, such as wired link
> as wireless it would be fairly easy to inject forged packets, from
> forged mac addresses etc...

I'm referring to the ability with some routing protocols that allow you 
to use a key to authenicate the data inside the protocol (such as ospf). 
But I agree with you that this is a seperate problem area.

> 
> Internet has the added benefit most backbone links are physically
> secure, such as fibre, or proprietary/secured microwave links, however
> with wireless the backbone is most likely to be made up by the end
> users, so it's a catch 22...

Yeah, though I doubt all that many people go around updating routers 
when problems have been disclosed to the public. I slightly recall a 
report thinger I read when people went looking and managed to compromise 
several "important" routers.

> 
> AG> (Yes, people can still cause problems if they compromise on of the boxes.)
> 
> What about links easily compromised?
> 

Well, its a different ball game. Depending on how good the attackers 
are, we may never notice.




To unsubscribe: send mail to majordomo at wireless.org.au
with "unsubscribe melbwireless" in the body of the message



More information about the Melbwireless mailing list