Abstract
Abstract The rapid advancement of edge devices has increased the volume of sensitive visual data. This creates a strong need to secure the data by incorporating efficient encryption algorithms. To overcome these challenges, this paper introduced a hybrid cellular automata (CA) with a chaotic map-based encryption scheme. The framework’s main goal is to guarantee the highest level of privacy in resource-constrained AIoT contexts. It uses hybrid cellular automata and rule vectors [204, 51, 51, 51, 195, 60, 60, 51]. It also uses the logistic and Tinkerbell maps to make the encryption process more complex and distributed, thereby increasing the overall security of the encrypted data against potential breaches. A dynamic S-box is generated using hybrid chaotic sequences to enhance nonlinearity and strengthen resistance to various attacks. The encryption framework developed a multi-layer permutation by using CA rules and chaotic shifting in the encryption process. Therefore, hybrid keys play a crucial role in pixel-level diffusion, maintaining high security with minimal computational complexity. Therefore, this algorithm is suitable for real-time use in AIoT edge devices. Security analysis demonstrates that the proposed framework achieves high entropy, low correlation, and robust resistance against different cryptographic attacks. This is proved by excellent NPCR, UACI, PSNR, and correlation coefficient values. The result indicates that the proposed algorithm provides strong security, efficiency, and privacy-preserving image encryption for next-generation AIoT applications.