Exploiting Compressed Sensing and DNA Coding for Multi-image Encryption Scheme With Application to Industrial Internet of Things

The rapid proliferation and extensive application of image data have made image encryption technology increasingly important in the field of information security. This article reports a novel multi-image encryption scheme for industrial internet of things (IIoT) applications. First, a hyperchaotic random number generation model (3-D hyperchaotic composed of Logistic map, Sine map and Linear functions (3-D-LSL)) is built using linear function, Sine map, and Logistic map. The model has high Lyapunov exponents and extremely wide state space, and can be implemented in hardware. After that, we devise a multi-image encryption algorithm (LSL-MIEA) based on this. The algorithm employs compressed sensing to handle the input images to decrease the volume of encrypted data. Subsequently, it utilizes DNA coding and bit-level diffusion to integrate the information of multiple images and fully achieve the confusion of pixel values. The simulation results demonstrate that the proposed scheme is highly safe, especially the values of number of pixel change rate (NPCR) and unified average changing intensity (UACI) reach 99.6079% and 33.4610%. It holds advantages in resisting differential attack and eliminating correlation when compared with existing algorithms. Finally, the application of the proposed scheme in IIoT is discussed, offering a solution for information protection within the IIoT environment.