[olsr-dev] Implementing an attack into olsrd

Sven-Ola Tuecke (spam-protected)
Mon Jun 19 10:42:23 CEST 2006


Hi,

using Pub/Priv keypairs is not really an option:

- U need the openssl/matrixssl lib installed (>1Mb footprint)

- Much cpu overhead to calculate keys each time you receive
  a message (no hardware enc in the WRT), this will kill any
  embedded box > 2 neighbours.

- Need to fiddle with all neighbours to fetch their pubkeys ->
  mesh network is not open any more / very hard to handle
  the org. Need to setup an Key-Expire-Infrastructure too.

I think it is more effective, to spot the location of a new
vandal and make a personal visit with a couple of mesh
members. In most cases this will be more easy than on the
internet - Wifi range isn't too large...

Sven-Ola

"Aaron Kaplan" <(spam-protected)> schrieb im Newsbeitrag 
news:(spam-protected)
> >
>> Afaik this is simply a shared key, which is not very secure in a 
>> community
>> network with several nodes and several different admins (in most  cases 
>> at
>> least > 50).
>>
> Yes that is how i remember the secure plugin.
> I don't know if we really want a signature checking (RSA, public key 
> crypto etc)
> in small embedded devices but anyway  - for a centralized approach - 
> signing the olsr packets with a cert (packet originator) which is  signed 
> by the master cert - is more what I was thinking about.
>
> There are distributed approaches as well. I can find out references  to 
> papers if needed.
>
>
>
>
>> though the threshold might be higher
>>
>> Jens
>> _______________________________________________
>> olsr-dev mailing list
>> (spam-protected)
>> https://www.olsr.org/mailman/listinfo/olsr-dev
>
>
> _______________________________________________
> olsr-dev mailing list
> (spam-protected)
> https://www.olsr.org/mailman/listinfo/olsr-dev 





More information about the Olsr-dev mailing list