Plaintext-Related Image Encryption And Its Application In Zero-Trust Environment

Digital image information as a visual way of communication is playing a more and more important role.However,the openness and sharing of the network make the image data vulnerable to illegal access and leakage in the process of use and circulation.Traditional encryption algorithms such as RSA,AES,ECC,etc.,if used to encrypt images,will have limitations due to large image size and strong correlation between pixels.Since chaos theory has been found to be suitable for image encryption,chaotic encryption has gradually become a mainstream technology in the field of image encryption.So far,the classical chaotic image encryption algorithms with a "scrambling-diffusion" structure using a fixed key stream still have security weaknesses that are vulnerable to selected plaintext attacks..In order to enhance the security of the cryptosystem,on the one hand,it can increase the number of encryption execution rounds,but it will reduce the efficiency;on the other hand,it can use plaintext related encryption scheme to improve the security and sensitivity,but it will bring additional key overhead,additional storage overhead and the limitation of being unable to resist arbitrary location clipping and noise attacks,which is difficult to meet some specific practical needs.In practical application scenarios,the zero trust principle advocated in recent years is an effective method to ensure data security in cloud storage services.However,the traditional encryption algorithm is still used as the main means in the cloud storage of image data,and the needs of application scenarios are not considered.The main research work of this paper is as follows:(1)A plaintext correlation encryption scheme based on LSB self quantization is proposed.The scheme generates plaintext related key stream by quantizing the disturbance value of plaintext image and applies it to encryption and decryption,which effectively enhances the ability to resist the chosen plaintext(cipher)attack.At the same time,the design of plaintext LSB repeated embedding disturbance mode,combined with the XOR control in the substitution stage of encryption design,and the mobility characteristics in the two scrambling sub stages,makes it possible to adaptively extract disturbance information and generate decryption key stream.Moreover,it overcomes the problems of existing plaintext related encryption mechanisms,such as the extra key cost,the storage cost exceeding the image itself,and the inability to resist clipping or noise attacks at any position.The experimental results show that the algorithm has the advantages of security,efficiency and practical performance.(2)In the zero trust image cloud storage scenario,the image encryption algorithm based on LSB self-quantization is integrated,and the parallel implementation of encryption and decryption algorithm is provided.By LSB self-quantization,the ability of the cryptosystem to resist chosen plaintext and chosen ciphertext attacks is effectively enhanced.By taking the row or column of the image as the processing unit,the whole algorithm process,including LSB repeated embedding and extraction,is executed in parallel.Experiments show that the algorithm can take security into account at the same time.Especially when encrypting a large number of images,parallel mode can also ensure real-time performance.It further shows that the algorithm has great application potential in the image cloud storage scene based on zero trust.