Statistical analysis of S-Box in Rijndael-AES algorithm and formulation of an enhanced S-Box

Abstract

The internet is widely used to support services ranging from education, medicine, entertainment to e-commerce. Cryptography helps to safeguard this transmitted information using authentication, digital signatures and encryption algorithms. In this paper we focus on AES, a symmetric block cipher algorithm which implements the Rijndael algorithm. We present a statistical analysis of the Rijndael-AES S-Box so as to evaluate the weaknesses present in the S-Box. These tests evaluate susceptibility of the AES S-Box to algebraic and statistical attacks. Using the obtained results, a technique of formulating a more non-linear S-Box is suggested. This technique uses the incursive congruential method, which produces highly non-linear output with a lower degree of correlation than the current AES S-Box.