Research On Chaos-based Image Coding&Encryption Algorithm And Hash Function Construction

With the development of Internet and wireless technology, it becomes very convenient,fast and real-time to transmit and duplicate image, voice, video and other multimedia datathrough a variety of mobile social network systems. In order to protect privacy, cryptologyhas become a hot research area, where encryption algorithm and Hash function are two basictechnologies, and play an important role to ensure the information security.Due to its track ergodicity, unpredictability and the sensitive dependence on initialconditions or control parameters, chaos theory based on nonlinear science is widely used inthe fields of control theory, signal processing, communication, computer cryptography etc.Among them, chaotic cryptology is an important research branch. The study of nonlinearscience and information science promotes the further development of chaotic cryptology sothat lots of research findings have been achieved. However, current approaches have theweaknesses in terms of safety and efficiency as chaos theory has not been completelyinvestigated at present. Based on the current research results of chaotic cryptology, this thesisis to investigate information security strategies based on digital chaos and coding theory,including the design of image coding encryption algorithm and the construction of Hashfunction based on chaos theory. Specifically, the main contributions and achievements of thisdissertation are as follows:1. Current research findings and advantages of chaotic cryptology are summarized,analyzed and reviewed, in particular focused on image encryption algorithm and constructionof Hash function based on chaos theory. The weaknesses of current approaches and the aimsof the research are also outlined.2. There are lots of periodic windows and its iteration sequences appear unevendistribution in Logistic mapping attractor. To enhance the ability, a spatial chaos-based imageencryption algorithm is proposed using the advantages of coupled map lattice model, such assimple structure and complex dynamic characteristics. The cryptosystem gains confusion anddiffusion by the coupling effect among coupled map lattices. In addition, the relation to thekey and plaintext can effectively resist known/chosen plaintext attack. Experimentalsimulation results show that the algorithm is of high safety.3. On the basis of the concept and properties of the ergodic matrix, a block-based imageencryption algorithm is proposed using the complex dynamics characteristic of spatiotemporalchaos. Firstly, the pseudorandom matrix is generated by iterating spatiotemporal chaos, whichobtains good randomness and cryptographic properties. Secondly, each image sub-block is diffused and confused through algebraic and scrambling operations using standard ergodicmatrices. Especially, the key is closely related to plaintext, and it makes encryption schemesimilar to “once a secret” cryptosystem, which improves the security of the encryptionalgorithm. Theoretical analysis and computer simulation results show that the proposedscheme is extremely sensitive to the key and plaintext, and can satisfy all kinds of existingimage encryption security requirements.4. According to the characteristics of JPEG image, combined with the JPEG codingprinciple, a chaos-based JPEG image encryption algorithm is proposed. The information ofimage is mainly focused on DC coefficients after discrete cosine transform, so only DCcoefficients are extracted and further encrypted, then the encrypted DCT coefficient is entropycoded and consequently the transmission data flow is generated. Theoretical analysis andexperimental simulation results show that the algorithm not only remains compression rateunchanged but also securely encrypts the image information.5. By using time and frequency domain properties of wavelet analysis, two imageencryption algorithms based on wavelet transform and chaotic sequence are proposed. Thebasic idea is to decompose original image into low and high frequency sub-bands, and the lowfrequency sub-band contains basic information of original image. Based on this characteristic,only the lowest frequency sub-band is encrypted to increase the encryption robustness andeffectiveness. Additionally, key stream is closely related to plaintext/cipher-text to improvethe ability to resist traditional attacks. Moreover, initial values and parameters of chaossystem are disturbed, which make the algorithm have better diffusion effect. Simulationexperiments verified that both two algorithms are safe and real-time, and have goodrobustness to common cracking attack.6. Studied on the statistical features of pseudo-random sequence generated fromfractional order chaotic systems, the results show that the statistical features satisfy theGolomb’s principles. So the fractional order chaotic system is suitable for image encryptionapplication. Taking fractional order Chen system as an example, experimental simulationresults verified the encryption scheme is safe and effective.7. According to the traditional definition and general structure of Hash function, a newspatiotemporal chaos-based Hash function construction method is proposed. Compared withthe existing chaotic Hash functions, the proposed method improves computationalperformance of Hash system. The theoretical analysis and experimental simulation prove thatthe method meets the Hash function requirements of invasive, forgery prevention,irreversibility and key sensitivity. Moreover, the maneuverability of the system is strong, and can be applied to industrial applications. In addition, the hardware implementation of thecorresponding HASH coding system is designed based on FPGA platform. It is applied to theSmart Power Grids for message authentication in the Integrated Meter Reading System.Experimental results show that this system has a high computing speed, can completemessage authentication quickly even for frequent requests, occupy minimal hardwareresources, and effectively strengthen the security of message transmission.