Informatica

Commonly modern symmetric encryption schemes (e.g. AES) use rather simple actions repeated many times by defining several rounds to calculate the ciphertext. An idea we previously offered was to trade these multiple repeats for one non-linear operation. Recently we proposed a perfectly secure symmetric encryption scheme based on the matrix power function (MPF). However, the platform group we used was commuting. In this paper, we use a non-commuting group whose cardinality is a power of 2 as a platform for MPF. Due to the convenient cardinality value, our scheme is more suitable for practical implementation. Moreover, due to the non-commuting nature of the platform group, some “natural” constraints on the power matrices arise. We think that this fact complicates the cryptanalysis of our proposal. We demonstrate that the newly defined symmetric cipher possesses are perfectly secure as they were previously done for the commuting platform group. Furthermore, we show that the same secret key can be used multiple times to encrypt several plaintexts without loss of security. Relying on the proven properties we construct the cipher block chaining mode of the initial cipher and show that it can withstand an adaptive chosen plaintext attack.

In our previous paper we presented an offline e-cash system with observers. We have shown that the proposed system satisfies basic requirements for e-cash schemes. We also covered such security issues as chosen message attack resistance and forgery of protocols data. However, in that paper we focused more on the system itself, rather than its analysis.

In this paper we consider an improved version of earlier published asymmetric encryption protocol based on matrix power function (MPF). Recently, a linear algebra attack on earlier version of this protocol was found. This attack allows an attacker to break suggested protocol in polynomial time.

New asymmetric cipher based on matrix power function is presented. Cipher belongs to the class of recently intensively evolving non-commuting cryptography due to expectation of its resistance to potential quantum cryptanalysis.