"Keith Packard" <[email protected]> writes:

> I think it would be sufficient to just add a 64-bit random nonce to the
> packets and leave the rest of the protocol alone.

Oops. For this to work, the pad boxes would have to remember the nonces
seen before and ignore them.

A longer time stamp in both directions should work though -- the ARM
command requires that the LCO compute the expected receive time of the
packet in terms of the PAD box time base; this two-way communication
avoids replay attacks of the whole protocol

        LCO                     PAD
t1      QUERY ->
                             <- REPLY
                                pad_tick
t2      ARM ->
        time = pad_tick + (t2-t1)
                                check time within 1/10 second of current time


We'll need a source of entropy for both LCO and PAD ends so that their
'clocks' start at a random time rather than zero. Both ends have a radio
chip; I'll see if I can't cook up some way to just get a packet of white
noise at startup time and pull some entropy out of that.

-- 
-keith

Attachment: signature.asc
Description: PGP signature

_______________________________________________
altusmetrum mailing list
[email protected]
http://lists.gag.com/mailman/listinfo/altusmetrum

Reply via email to