基于物理随机位生成器与混沌像素交叉互换的图像加密算法

目的为了解决当前混沌图像加密技术忽略了随机序列产生的时间延迟现象,且难以克服其自身迭代的周期性,使其序列的自相关性不理想,导致密文安全性不佳等问题。方法引入级联耦合混沌半导体环形激光器,设计基于物理随机位生成器与混沌像素交叉互换的图像加密算法。首先引入SHA-256散列函数,利用明文像素值,生成一个256位的密钥,并将其分割为一系列的8位子密钥;利用这些子密钥来计算Logistic-Sine复合映射的初始条件,以输出一组随机序列;根据混沌序列,定义像素交叉互换机制,对输入明文进行预处理,消除相邻像素之间的相关性;基于级联耦合混沌半导体环形激光器,充分利用其自身的时间延迟与交叉反馈的特性,设计物理随机位生成器,以同步输出考虑时间延迟的控制矩阵与随机位流;将Logistic-Sine复合映射输出的混沌序列转换为一个过渡矩阵,联合控制矩阵,定义像素混淆机制,彻底改变明文的像素位置;最后,利用随机位流,设计像素联系扩散函数,改变图像的像素值。结果实验结果显示,与当前混沌加密技术相比,所提算法具有更高的安全性与鲁棒性,能够有效抗击明文攻击,相应的密文熵值约为7.9958,且NPCR(7)Number of Pixel Change Rate(8)、UACI(Unified Average Changing Intensity)分别为99.50%、33.46%。结论所提加密算法具有较高的安全性和抗攻击能力,能够安全保护图像在网络中传输,在信息防伪等领域具有较好的应用价值。

Image Encryption Algorithm Based on Physical Random Bit Generator and Chaotic Pixel Cross Interchange

The work aims to solve the defects as unsatisfactory sequence autocorrelation that causes low cipher security induced by neglecting time delay generated by the random sequences in current chaotic image encryption technology and difficulty in overcoming the periodicity of its own iteration. The cascade-coupled chaotic semiconductor ring laser was introduced to design an image encryption algorithm based on physical random bit generator and chaotic pixel cross interchange. Firstly, the SHA-256 hash function was introduced and a 256-bit key was generated with the value of the plaintext pixels, and then the 256-bit key was divided into a series of 8-bit sub-keys. A group of chaotic sequences were outputted by calculating the initial conditions of Logistic-Sine composite map with these sub-keys. Then, the pixel cross interchange mechanism was defined according to the chaotic sequence and the input plaintext was preprocessed to eliminate the correlation between adjacent pixels. The physical random bit generator was designed based on the cascade-coupled chaotic semiconductor ring laser which made full use of its own time delay and cross feedback to synchronously output the control matrix and stochastic bit stream that considered the time delay. The pixel confusion mechanism was defined by transforming the chaotic sequence outputted by the Logistic-Sine composite map into a transition matrix and jointing the control matrix to completely change the location of plaintext pixels. Finally, the pixel diffusion function was designed with random bit stream to change the pixel value of the image. The experimental results showed that the proposed algorithm had higher security and robustness which could effectively resist against the plaintext attack compared with the current chaotic encryption technology, and its corresponding cipher entropy was about 7.9958, and the values of NPCR (Number of Pixels Change Rate) and UACI (Unified Average Changing Intensity) were 99.50% and 33.46%, respectively. With higher security and anti-attack capability, the proposed encryption algorithm can protect the safe transmission of images in the network and it has better application value in the field of information security and anti-counterfeiting.