A chaotic-system-based parallel image encryption algorithm with orthogonal arrays supporting thumbnail decryption

Abstract

Cloud computing is a technology which offers users with on-demand access to various resources, and the protection of user information has turned into one significant challenge for cloud service providers (CSPs). In order to enhance the security of image files while meeting the performance demands of cloud computing, this study mainly focuses on the protection of image information in cloud environments, and then an efficient multi-image encryption strategy with fusion of DNA coding and orthogonal arrays is proposed. This scheme is tailored for parallel computing environments in the cloud, which can be able to achieve multi-threaded encryption/decryption for large numbers of images with various sizes at high speed. During the process of decryption, it allows users to spend less time decrypting the encrypted images into thumbnails or selectively decrypting specific images for easier browsing, which significantly reduces data transmission and server computation loads. The encryption keys are generated through coupling one discrete memristor and a sine map into one new chaotic map, then the orthogonal arrays and DNA coding for bit-level encryption and permutation of images are supplied to enhance the encryption security. Simulation tests and security analysis indicate that the strategy achieves high levels of safety and reliability while delivering excellent performance. The more images being encrypted, the higher the encryption efficiency will be. In the test environment, the speed can reach up to 24 MB/s, which is far exceeding encryption algorithms implemented on the same platform. In addition, this study further extends the application of proposed encryption scheme to the encryption of video data, the relevant analysis and simulations are supplied to verify the feasibility and effectiveness of the scheme.