A Multiple-Image Encryption Algorithm Based on Novel 2D Memristive Hyperchaos and Virtual Rubik’s Cube Transformation

With the explosive growth of multimedia data, multiple-image encryption that ensures both high security and high efficiency has become a key research focus. Existing algorithms often fail to balance security and efficiency, thus highlighting the need for a scheme that optimizes both. A new 3D spatial multiple-image encryption algorithm based on hyperchaos is proposed to enhance the security and efficiency of digital images in this paper. First, the properties of the discrete memristor are used to create a cross-coupled hyperchaotic map. The newly created chaotic map performs better chaotically than existing two-dimensional chaotic maps with the sample entropy, Lyapunov exponent, bifurcation diagram, and other factors. Second, a new virtual Rubik’s cube transformation is designed to change the pixel positions. The random size and random direction of the transformation overcome the limitation of the fixed Rubik’s cube size in traditional transformations. This increases the spatial complexity of the scrambling process and further enhances the security of the encryption. Finally, a 3D cross-indexed diffusion model is constructed to diffuse each pixel value across the entire data cube via 3D random coordinates, thereby generating the encrypted images. Performance analysis shows that the average encryption time is 0.28 sec for [Formula: see text] images and the information entropy of the encrypted images approaches 8. The results of the analysis demonstrate the excellent efficiency and extremely high security of the suggested multiple-image encryption algorithm.