AI-Guided Permutation–Diffusion Framework for Chaotic Image Encryption
International Research Journal of Modernization in Engineering Technology and Science · 2025
The rapid growth of multimedia communication has raised significant challenges in securing digital images against unauthorized access and malicious attacks.Conventional image encryption techniques based on chaotic maps, permutation-diffusion models, or symmetric key cryptography often exhibit vulnerabilities when deployed in high-dimensional, dynamic environments.This paper introduces an AI-guided image encryption framework that leverages permutation strategies enhanced by intelligent key scheduling.The proposed approach integrates chaos-based permutation with machine learning-assisted optimization, ensuring improved randomness, high entropy, and resilience against differential and statistical attacks.Experimental evaluations demonstrate near-ideal values of NPCR (99.61%) and UACI (33.45%), along with entropy close to 7.999, validating robustness.Comparative analysis with state-of-the-art literature confirms that the AI-guided permutation approach achieves superior resistance to brute-force and correlation attacks while maintaining lightweight computation suitable for IoT and cloud platforms.The modular design further allows adaptability to diverse image formats, making it a scalable solution for next-generation secure image communication systems.