Substitution box design and image encryption application using operators of genetic algorithm based on evolutionary computation

This study proposes a method that employs the crossover and mutation operators of real-coded genetic algorithms iteratively to design a high-security S-Box that meets specific security requirements. The relevant operators of genetic algorithms enable the exploration of the solution space efficiently, facilitating the discovery of optimal S-Box structures. As a result, both computational efficiency is improved, and the most suitable solutions that satisfy security requirements are obtained. The designed S-Box has been evaluated in terms of performance using various security criteria and has demonstrated effective results when compared to existing S-Box designs. Conducted tests reveal that this S-Box provides lower error rates, higher complexity, and better security in encryption and decryption processes. Additionally, the developed S-Box has been integrated with the 2D hyperchaotic Styblinski-Tang map to propose a highly secure method in the field of image encryption. This integration ensures a high level of confusion and security in encrypting image data, as both systems exhibit highly chaotic properties, making the encryption process difficult to break. The security of the designed encryption method has been supported by comprehensive analyses. Statistical evaluations confirm that encrypted data should not exhibit any unexpected patterns, meaning there should be no vulnerabilities or detectable structures. Furthermore, the encryption method has been tested for resilience against various attacks and external noise. These robustness tests demonstrate how effective and secure the encryption algorithm is under both theoretical and practical conditions. In conclusion, the developed method not only provides high security but also holds broad application potential, offering a practical and effective solution for various fields, such as image encryption.