A novel block substitution technique for image/text encryption using memory loss models and fuzzy mutation

Robust block substitution techniques are essential to protect information against sophisticated cryptographic attacks. Existing techniques use static methods, mainly Sbox substitution, to increase nonlinearity. These static substitutions have significant weaknesses, however. Bit alterations impacts only the specific block where the bit alterations occur without extending their effects to subsequent blocks. Second, current block substitution techniques only leverage alterations in input symbols to boost confusion without considering other significant changes such as the positions of the altered symbols within the input block. To address the limitations, a more dynamic block substitution is presented. This new block substitution technique effectively monitors changes in input symbols and extends the impact of these changes to cover the entire ciphertext. The proposed method utilizes operations based on chaos, deep symbol mutation, and memory loss models to process the input blocks. We analyzed the performance of the technique using robust testing tools. Our experiments demonstrated that the proposed technique surpasses existing methods in nonlinearity and randomness metrics, achieving SAC (0.50190), BIC (111.67), NL (112.0), and entropy (7.99997). Furthermore, we integrated the proposed substitution technique into both text and image encryption techniques. The experimental study showed a substantial improvement in the performance of these techniques when using the proposed substitution technique.