Re: Regarding use of Reed-Solomon code in wireless networks

Hi Detlef, Khaled,

Thanks for sharing your views.

For the detection mechanism to work, retransmission must be explicitly
switched off (as retransmitting same nonce gives advantage to the wormhole
attacker).

One way to avoid actually measuring the delay is to stipulate that the
response nonce from the receiver must come back after a SIFS delay as it
happens for any ACK frame in IEEE 802.11. This is possible if minimal
processing is involved at the receiver node.

If the response doesn't come back after SIFS, the receiver is probably far
away and is likely to be connected by a wormhole to this sender. That way,
we should be able to flag at least long wormholes.

Do you think the above heuristic will work in real networks?

Now if the nonces are encoded with RS code, is it still possible for the
receiver to decode the nonce, process it, encode it back with RS code and
transmit it after a SIFS delay from the time of reception?

Thank you in advance for your valuable comments.




Regards,
Debarshi Kumar Sanyal
KIIT University, Bhubaneswar



On 6 May 2015 at 17:47, Khaled Elsayed <[email protected]> wrote:

> Yes, it is usually a good idea to use RS encoding to combat channel
> problems.
> The time taken usually is function of the RS code word length and the
> parity bits.
> Usually it is very fast compared to other blocks like Viterbi decoder etc
> if you are talking about lower layers where Viterbi is common.
>
> On Wed, May 6, 2015 at 8:14 AM, Debarshi Sanyal <[email protected]>
> wrote:
>
>> Hi,
>>
>> Hi,
>>
>> We were working on design of wormhole detection methods in MANETs.
>> To achieve detection, we propose to measure the time taken for a test
>> nonce
>> to move from one node to it's neighbor node. If the time taken is larger
>> than the expected time for a packet to travel between two neighbor nodes,
>> the link is probably a wormhole link.
>>
>> Now to combat channel errors, is it worth encoding the nonce with
>> Reed-Solomon code?
>>
>> Our understanding is that propagation time between neighbor nodes in
>> commodity wi-fi setups is much smaller than the time taken to encode and
>> decode a nonce (say, 64 bytes long). So delay variations in encode/decode
>> process will easily mask any delay in propagation time.
>>
>> We would be immensely thankful if you could throw some light on this since
>> we do not have access to hardware platforms to get real measurements. We
>> are interested to know the approximate encode and decode times for RS code
>> on common hardware platforms.
>>
>>
>>
>>
>> Regards,
>> Debarshi Kumar Sanyal
>> KIIT University, Bhubaneswar, India
>> _______________________________________________
>> end2end-interest mailing list
>> [email protected]
>> http://mailman.postel.org/mailman/listinfo/end2end-interest
>> Contact [email protected] for assistance.
>>
>
>
_______________________________________________
end2end-interest mailing list
[email protected]
http://mailman.postel.org/mailman/listinfo/end2end-interest
Contact [email protected] for assistance.