Duple Color Image Encryption System Based on 3-D Nonequilateral Arnold Transform for IIoT

Abstract

The Industrial Internet of Things (IIoT) is to continuously integrate network communication, intelligent analysis and other technologies into every link of the industrial production process, thus greatly improving the manufacturing efficiency. While such systems greatly improve the productivity of the industry, they introduce major safety challenges during the design and operational phases. Therefore, it is necessary to encrypt and protect the information transmitted by the system during the design and runtime stages. This paper focuses on the transmission security of image data in every link of IIoT. Two color images that need to be transmitted can be encrypted into color ciphertext image by using 3D Non-equilateral Arnold Transform (3D-NEAT) and 3D Lorenz System (3D-LS). Different from the traditional algorithm which encrypts one plain image into one ciphertext image, our algorithm encrypts two color images into one color ciphertext image which will cause great confusion to the attacker whom illegally breaks the ciphertext image. First, the proposed method converts two color images with <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$N\times M$</tex-math></inline-formula> into a 3D bit-level matrix with size <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$N\times M\times 48$</tex-math></inline-formula> . Next, 3D-NEAT is applied to permutate the positions of the elements of the resulted 3D bit-level matrix. Then, the permutated 3D bit-level matrix is transformed into three 2D pixel-level images and then diffused by the random diffusion sequences which 3D-LS generates. Finally, the scrambling matrices generated by 3D-LS are used to scramble three diffused 2D pixel-level images and the output is considered as three color components of the encrypted image. The numerical experiments and security analysis show that the proposed image encryption scheme has strong resistance to several known attacks, and yields near-zero correlation and near-eight entropy for the RGB cipher image, and some performance is better than some of the recently proposed image encryption algorithms.