Enhanced double random phase encryption with logistic map keying and iterative diffusion
Mahmoud Karmany, Azza Moawad, Mohamed Abaza, A. Afalou, A. Mansour · Applied Optics · 2025
The critical requirement for secure image transmission in areas such as surveillance, medicine, and digital forensics has increased the importance of visual information security. Double random phase encryption (DRPE) has been extensively applied in optical image encryption owing to its feasibility in Fourier optics and its capability to encrypt effectively in both the spatial and frequency domains. However, despite these strengths, traditional DRPE techniques have serious drawbacks, including a limited key space, poor resistance to statistical attacks, and vulnerability to cryptographic exploration and known-plaintext attacks owing to the compromise of phase masks. To address these issues, this study proposes an enhanced DRPE (EDRPE) system that combines chaotic key generation based on a logistic map with a dynamic substitution box (S-Box) that offers enhanced confusion and diffusion. Additionally, to address the possible Grover's attacks on image ciphers, we propose a feature-iterative XOR image encryption (FIXIE) algorithm that generates image-dependent keys across multiple rounds. This new method, to the best of our knowledge, overcomes the inefficacy of classical DRPE by introducing a nonlinear, chaos-driven iterative key evolution model, achieving improved security without compromising computational efficiency. The simulation results indicate that the proposed EDRPE-FIXIE framework achieves strong unified average change intensity values, a high number of pixel change rates, extremely weak correlation between adjacent encrypted pixels, and close-to-ideal entropy. EDRPE-FIXIE fulfills the requirements for real-time and memory-restricted encryption services because these improvements are achieved without compromising memory or processing speed.