An enhanced method for constructing cryptographically strong dynamic S-boxes with high and invariant nonlinearity via extended affine transformations
摘要
The rapid development of digital technologies and the Internet is creating new problems in the field of information security. Consequently, the demand for encryption algorithms that ensure secure communication between electronic devices is increasing on a daily basis. There is an urgent need to improve the effectiveness of symmetric key encryption algorithms to ensure the confidentiality and integrity of wired and wireless communications. The cryptographic strength of symmetric key encryption algorithms is contingent upon the substitution box (S-box), which is the primary component of these algorithms. Static S-boxes are employed in the encryption standards of numerous nations. The primary advantage of these methods lies in their simplicity and efficiency; however, their fundamental weakness lies in their comparatively limited stability when contrasted with advanced differential and linear analysis methods. Dynamic S-boxes are known to undergo modifications depending on the key or text during the encryption process. This approach enhances the cryptographic strength of the algorithm and substantially reduces the probability of prediction by an attacker. The present article addresses the issue of generating dynamic S-boxes with robust cryptographic characteristics for symmetric key encryption algorithms. The calculation of dynamic S-boxes is carried out using affine transformation. The reliability of the obtained dynamic S-boxes was assessed in terms of nonlinearity, differentiability and other cryptographic criteria. The proposed method demonstrates the feasibility of constructing dynamic S-boxes that are resistant to a variety of cryptographic requirements. This can be achieved by selecting elements from a directed adjacency matrix. The research has demonstrated that all dynamic S-boxes created by this method exhibit a minimal nonlinearity of 112 with a differential convergence probability of 4/256 and an average strict avalanche criterion (SAC) of approximately 0.5. The results of this study show that the proposed method is better than other similar dynamic S-box computation algorithms when it comes to resisting several main cryptographic attacks. This approach paves the way for further development of dynamic S-box design methods and provides valuable theoretical and practical insights into significantly increasing the attack robustness of modern cryptosystems.