Hey all, waiting for a new computer to install my gtx260 (older was not
powerfull enough and overheat :-s) im going into the theoritical stuff.
I have some questions. The project is using rainbow tables. Is this the version
presented by Philippe Oechslin? Or is it just the standard time-memory tradoff
from Hellman?From what i know about rainbow table, it is only a big table with
a reduction function different for each column of the table. Advantages of
rainbow table is that merging can only occurs at the same level due to the
different reduction function between level (by level i mean column of the
table).Now for this project, we are talking of generating 2^(8.5) tables. So
this is not a unique rainbow table right? What is the goal of having more than
one table if we use so-called rainbow table? Can anyone explain to me?Maybe the
goal is to decrease the look-up time that is in theory N^(2/3) (N the keyspace
size) by having a smaller N value and get a distributed lookup within more than
one rainbow table? Can a1 confirm (or not)?
Another thing is about the RRCipherModeCompleteMessage. This message gives 4
frames of 114 bits, so why do we limit the size of the arrival space to 64 bits
messages? How is the mapping between RR message and 64-bit keystream done?Since
the whole 114-bit RR message is encrypted with the keystream output of A5/1
that is also 114bits, i don't understand why we are using a sliding version of
a 64-bits keystream to get 50 ciphertext messages.
Can anyone explain to me what i've missed?
Thanks to all.
Hopping to be able to start my computation soon^^cheerssylv1
_______________________________________________
A51 mailing list
[email protected]
http://lists.lists.reflextor.com/cgi-bin/mailman/listinfo/a51
