[olsr-dev] Selfish LQ

[giuseppe de marco]

Sven-Ola Tuecke wrote:
> Hi,
>
> No. Does not make sense. A DOS attack to a radio mesh is more easy to 
> deploy using standard 2.4ghz videolink devices and a big antenna. 
> Standard procedure to handle this:
>
> - add iptables to keep the node out of routing
> - if this does not help: track location, visit & change physical 
> conditions
>
> P.S. you may invent a mesh-of-trust technique. That's missing somehow...
>
> // Sven-Ola
>
> "giuseppe de marco" <(spam-protected)> schrieb im Newsbeitrag 
> news:(spam-protected)
>> Suppose I am a malicious node.
>> Suppose that the true LQ toward a node is 4, but I declare (forge) a 
>> HELLO packet in such a way the LQ toward that node
>> will be 1. In this way, possibly, a node will route its traffic 
>> through me and not through other legitimate nodes.
>> It makes sense or I am loosing something?
>> G
>>
>> -- 
>> Giuseppe De Marco, phD
>> Toyota Technological Institute
>> 468-8511 Aichi 2-12-1 Hisakata, Tenpaku-ku, Nagoya, Japan
>> Email: demarco at toyota-ti dot ac dot jp
>> Tel (int): +81 (052)-809-1802
>> Skype-Id: giuseppe_dem2
>>
>>
>> _______________________________________________
>> 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
>
Maybe I have beenn misunderstood or I'am not understanding.
The problem here is a follows:
     1              2        
1---------2-----------3
  \   1                  /
    \T=======/
          1
In this net, 1,2,3 are good, i.e. they say the true about LQ of their 
link. LQ=p_f * p_r
But, node T is malicious: he forge a high value for p_r on link T-3, in 
such a way the route from 1 to 3 has LQ=2, i.e. better than  1-2-3. Thus 
, OLSR will choose 1-T-3.
This has 2 conquences: 1) the traffic will pass through me 2) the route 
will be (perhaps) bad.
How can I change the iptables in T or 1 and 2 if  1 and 2 don't know 
that T is using a bad algorithm? 1 and 2 should have access to the links 
of T in order to judge if T is
saying the truth about the LQ...
Thanks

G

-- 

Giuseppe De Marco, phD
Toyota Technological Institute
468-8511 Aichi 2-12-1 Hisakata, Tenpaku-ku, 
Nagoya, Japan
Email: demarco at toyota-ti dot ac dot jp
Tel (int): +81 (052)-809-1802
Skype-Id: giuseppe_dem2