Ana Kukec escribió:
> Hi all,
>
> Just to summarize the possible options about encoding hash algorithm:
>   
right thanks

> a) Use the same hash algorithm as in CGA for all hashes. There is no 
> bidding down attack in that case. Different algorithms then CGA 
> algorithm for other hashes does not increase security.
>   

right
actually, as someone suggested in the meeting, it is also possible to 
encode the hash function in the last 3 bits of the address, 
irrespectivly the address is a CGA or not
> b) Use the Hash Algorithm option to define different (or the same?) 
> algorithm for all hashes. It is vulnerable to the bidding down attack, 
> but provides flexibility, since in the future, SeND might be used 
> without CGAs.
>
>   
well, the scope of bidding down attacks in this case could be limited by 
admin configuration. I mean, if it is determined that a given hash 
function is insecure, then hosts simply don't accept the hash algorithm 
any more and no bidding down attack is possible

The problem is when two hash algorithms with different level of security 
must be accepted simultaneously, so during that period, the attack will 
be possible

regards, marcelo



> Any opinions? :-)
>
> Ana
> _______________________________________________
> CGA-EXT mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/cga-ext
>
>   

_______________________________________________
CGA-EXT mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cga-ext

Reply via email to