Image Encryption Algorithms Based On Multi-Scroll Memristive Chaotic Systems

With the development of 5G,artificial intelligence,and the Internet of Things,digital images are widely transmitted on various communication networks.At the same time,the security of digital images is more and more widely concerned by countries,enterprises and individuals.Therefore,the design of secure and effective encryption algorithms for digital images has become a research hotspot.Chaotic system has many intrinsic properties such as sensitivity to initial state and unpredictability,which makes it an ideal choice for cryptography.Therefore,in this thesis,based on the Chua’s system,a multiscroll memristive chaotic system with rich dynamic behaviors is proposed.Secondly,based on the proposed multi-scroll memristive chaotic system,a new image encryption algorithm was designed.Finally,combined with compressive sensing technology,a color image encryption algorithm was proposed.The main research of thesis is as follows:1.A four-dimensional multi-scroll memristive chaotic system is proposed by using the magneto-controlled memristor model as a nonlinear term to introduce the classical Chua’s system.Through dynamic analysis,it is found that the system has complex dynamic behaviors: selecting and adjusting the internal control parameters of the memristor can generate any odd or even number of double scroll attractors.By adjusting the memristive coupling strength and the initial state variable,continuous robust chaotic sequences can be obtained.By adjusting the initial state variable,the coexistence attractors of the same structure in different positions can be seen,so the system has homogeneous multistability.2.A novel image encryption algorithm is designed based on the multi-scroll memristive chaotic system proposed in this thesis.The algorithm has the following three characteristics:(1)The hash value of the plaintext image is used in the generation and selection stage of the chaotic sequences.The chaotic sequences is involved in every link of permutation and diffusion,so the algorithm is highly sensitive to the plaintext image;(2)Knuth-Durstenfeld shuffling algorithm is used for two times of bit-level scrambling,which changes the pixel value while changing the pixel position,greatly increasing the randomness in the encryption process;(3)DNA dynamic coding and computation are adopted in the diffusion stage.Through performance analysis and comparison with other algorithms in recent years,it can be found that the algorithm can cope with various attacks,and shows good robustness and security.3.A color image encryption algorithm based on magnetized HNN multi-scroll memristive chaotic system is designed.The algorithm has the following two characteristics:(1)The hash value of the plaintext image is used as the external key to perturb the four internal keys and control parameters of the system,which expands the key space of the algorithm and improves the sensitivity of the algorithm to the plaintext.(2)In order to reduce the consumption of resources in the process of color image encryption,compressed sensing technology is introduced to compress the three channels of the color plaintext image and extract the main data for encryption.Experimental simulation and performance analysis show that the algorithm has a good reconstruction effect and security for different sizes of images with different compression rates.