Research On Digital Image Encryption Scheme Based On Discrete Memristive Chaotic System

With the continuous development and progress of the Internet and mobile communication devices,people can browse and exchange digital images faster.Due to privacy and commercial security,digital images transmitted through public channels need more secure and effective protection.Therefore,the security of digital images has received wide attention and become a research hotspot.Among the various protection techniques about digital images,image encryption is the most direct and effective one.Chaos image encryption is an important branch of multitudinous image encryption schemes,because chaos has the characteristics of aperiodicity,sensitivity to control parameters and initial values,and ergodicity,so that the pseudo-random sequence generated by the chaotic system meets the requirements of the password in the image encryption.In addition,the researchers found that the memristive chaotic system generated by coupling the memristors and the chaotic system has a more complex nonlinear dynamical behavior.As a class of memristors,discrete memristors can be coupled to discrete chaotic systems,which can produce super-chaotic phenomena under the initial value and parameter control,which provides a new idea for finding chaotic systems more suitable for image encryption.At present,the research and application of discrete memory resistance chaotic systems are not very extensive.therefore,Starting from the two aspects of chaotic systems and encryption schemes,this thesis makes a study on the design of novel discrete memristive chaotic systems and the design of encryption schemes with higher security.In addition,we also study digital image encryption based on non-adjacent parallelable permutation and dynamic DNA-levels two-way diffusion.The main studies of this article are as follows:Firstly,in order to improve the complexity of discrete chaotic systems,in this thesis,we propose a modified cosine discrete memristor,and coupled it to the Cubic map,to obtain a new two-dimensional discrete memristors chaotic system.By analyzing the new system through bifurcation diagram,its dynamical behavior is rich and it is in a superchaotic state under specific conditions.The time-series randomness test is verified that the chaotic sequences produced by this chaotic system have excellent stochastic properties and can be applied to image encryption.Secondly,combined with the proposed two-dimensional discrete memristive chaotic system,we design an image encryption scheme based on circular shift,bidirectional diffusion and global permutation.The encryption scheme adopts an iterative structure and arranges different permutation processes before and after the diffusion process.To further improve the efficiency of image encryption,a novel bidirectional diffusion strategy is adopted during the diffusion process.The analytical results of key space,histogram,correlation,key sensitivity,plaintext sensitivity,shear attack and information entropy analysis show that the proposed scheme has extremely high security and practicability.Finally,this thesis presents a novel image encryption scheme based on non-adjacent parallelable permutation and dynamic DNA-level two-way diffusion based on 2D Logistic-Sine-Coupling Map.The scheme logically divides the plaintext images into adjacent blocks and displacement blocks,where the displacement blocks are randomly arranged by multiple computational units.The displaced images were segmented and encoded into four DNA planes.When each DNA plane is encoded,the computing unit dynamically adopts the coding rule according to the chaotic sequence,combined with the dynamic DNA decoding,ensuring the excellent nonlinearity of the diffusion process.The hash value,the key component,and the chaos value jointly determine the DNA-level bidirectional diffusion method conducted in parallelable in the DNA plane.Thus,the minimal change in any pixel is effectively spread throughout the image.The proposed scheme demonstrates its excellent performance through extensive simulation experiments and theoretical analysis.