一种图像置乱算法的FPGA实现方案——以Arnold算法为例

The FPGA-Based Implementation Scheme for Image Shuffling Algorithm—Taking the Arnold Algorithm as an Example

在5G与物联网时代,研发国产原创性加密芯片是信息安全研究领域的重点问题.相当多的设备会生成图像数据,以便用户查看使用,常见的RSA算法、DCT域图像水印算法、Rossler混沌变换算法、Arnold变换算法可以实现软件层面的图像加密,但硬件化的图像加密芯片方案却寥寥无几.为保证用户数据安全,以图像加密中常用的置乱算法为例,在FPGA上实现任意方形尺寸图像的Arnold加密算法,为图像置乱算法的硬件化提供了借鉴思路.之前的经验证明:即使一些在软件上非常简单的算法,在硬件上实现时均需要大量的数字电路元件才能完成.重点是使用已经证实能够在软件方面有效加密的算法,尝试应用在硬件上,希望达到低成本的硬件加密效果.实验结果表明:当在FPGA实现Arnold变换时,执行速度比用电脑实现Arnold变换快很多倍.

In the era of 5G and IoT, the development of domestically produced encryption chips is a key issue in the field of information security research. A considerable number of devices generate image data for users to view and use. Common RSA algorithm, DCT domain image watermarking algorithm, Rossler chaotic transform algorithm, and Arnold transform algorithm can realize software level image encryption. However, there are few hardwareencrypted image encryption chip schemes. In order to ensure user data security, this paper takes the shuffling algorithm commonly used in image encryption as an example to realize arbitrary square size on FPGA. The Arnold encryption algorithm of the image provides a reference for the hardwareization of the image scrambling algorithm. Even if some algorithms in the software are very simple , in the hardware to implement, they all need a large number of digital circuit components to complete. This article focuses on the use of algorithms that have been proven to be effectively encrypted in software, trying to apply them to hardware in the hope of achieving low-cost hardware encryption. The experimental results of this paper show that when the FPGA transform is realized, the execution speed is many times faster than the speed of using the computer to realize Arnold transform.