Hello everyone, suppose there is such a DES: Ek3 (Ek2 (Ek1 (T))) - I had a question
(yes, stupid) - why for the same DES, but with two keys, an attack of the type

“a meeting in the middle” is possible, but not with three? After all, the principle of the attack with two keys - as I understand it - if DES is like this: DES: Ek2 (Ek1 (T)) - then the principle is like with a nested doll - I will go through all possible options for k2 with a length of 56 bits and write all this into a table call it m1, and at one point, I’ll go to k1, which is an element of k2, and start sorting out k1 - also listing the results in a table, let's call it m2. And then I simply compare the tables m1 and m2 and the coincidences of the elements - this is going to be a “meeting in the middle” - right? BUT, why I can not work on the same nesting principle with DES with three keys? - just creating a table m3 and comparing the coincidences between the three tables already ?!

    1 answer 1

    In addition to the simple Encounter-Middle algorithm , there is Multidimensional-Encounter-Middle . Which is just designed for a greater number of investments. It is Meeting-Middle that reduces the complexity of 3DES to 2 ^ 112. Although the expected complexity should be 2 ^ 168.