Image encryption based on module learning with error using dynamic S-boxes

Implementing robust cryptography techniques in this digital world is important to provide security against unauthorized access, interception, or tampering data while transmitting over an insecure channel. An encryption technique for enciphering confidential images using dynamically generated multiple substitute boxes from Module Learning with Error is proposed to enhance security. The initial hash value is obtained by taking the secret image as input to the Secure Hash Algorithm(SHA-512), followed by an iterative call on the SHA-512 using the previous hash values to obtain multiple hash values. These hash values are used to generate multiple substitute boxes. The substitute boxes are selected randomly using a random sequence generated from an enhanced Logistic map to perform confusion for each \(16 \times 16\) $16\times 16$ image block. A random matrix of dimensions equal to the confidential image is generated using multiple hash values. The substituted image is operated with XoR operation with the random matrix generating the cipher image. The initial hash value is securely shared using Lattice-based Module Learning with Error. Different statistical and security analyses are carried out on the generated substitute boxes and cipher images of the proposed method. The different tests demonstrate that the proposed image encryption has a large keyspace, high encryption efficiency, and the ability to withstand different cryptanalytic attacks effectively.