The symmetric block cipher proposed in our work, unlike the currently widely used ciphers, is based on a new concept and does not require any additional rounds. Therefore, it has some similarities with the one-time pad technique, which up to now was the only perfectly secure cipher. However, a trivial method for breaking this cipher is known if an attacker manages to obtain two pairs of plaintext and ciphertext encrypted using the same key. This fundamental flaw prevents any implementations of the one-time pad technique in various encryption modes. For this reason, this technique is often viewed as a theoretical concept, but its practical applicability is highly restricted. Hence, one of our tasks is to show that the proposed cipher does not have this drawback and can be used to create symmetric encryption modes. We currently use the matrix function together with a non-commutative group to define the CTR mode of the block cipher. Interestingly enough, the application of the non-commutative group implies additional restrictions on the secret key, which aggravates the cryptanalysis of our cipher.
Project results:
Investigation of the avalanche effect and the bit independence criterion (BIC) for the proposed symmetric cipher was performed. Based on the early results, we proposed some enhancements to our symmetric cipher, which improved the previously mentioned statistics. To evaluate the advantages of our scheme, we performed a comparison, where the values of the avalanche effect and BIC for different parameter values of our block cipher were compared with AES128, DES and TDES algorithms. During the project, we found that additional scenarios could be tested, i.e. when the matrix Y has no constraints and scenarios when an additional mapping is introduced to improve mixing. Also, together with other members of the research team, resistance against passive attacks has been proven.
Period of project implementation: 2022-07-01 - 2022-08-31
Project coordinator: Kaunas University of Technology
