一种面向硬件实现的模数混合混沌真随机数发生器
Hardware-Oriented Modular Hybrid Chaotic True Random Number Generator
解旭辉 1胡汉平 2郑俊 3曾宇1
作者信息
- 1. 华中科技大学 人工智能与自动化学院,武汉 430074
- 2. 华中科技大学 人工智能与自动化学院,武汉 430074;图像信息处理与智能控制教育部重点实验室,武汉 430074
- 3. 华中科技大学 网络空间安全学院,武汉 430074
- 折叠
摘要
在信息安全领域,随机数发生器的应用越来越广泛.混沌的类随机性和不可预测性可以被用来设计性能良好的随机数发生器,然而,当混沌系统在数字电路实现时,会出现退化,存在严重的安全问题.本文利用模数混合混沌模型设计和实现了低成本、高效率的真随机数发生器,通过用两级比较器代替 ADC实现混沌量化获得真随机比特以减少资源消耗,以易于硬件实现的梳状混沌映射为基本单元构造了超混沌系统,模拟混沌系统采用光混沌方程实现简单且可以获得更高的速率.利用 FPGA 和 FPAA 对真随机数发生器进行联合硬件实现,整体电路简单,以相对较低的资源开销产生的真随机序列具有高稳定性和优良的统计性能,通过了 NIST SP800 标准随机性测试以及最为严格的 TestU01 测试.该设计有望集成到专用芯片中并广泛用于网络安全的各个领域.
Abstract
In the field of information security,the application of random number generator is more and more extensive.The quasi-randomness and unpredictability of chaos can be used to design a random number generator with sound performance.Nevertheless,when the chaotic system is imple-mented in digital circuit,it will degrade and there will be serious security problems.In this study,a low-cost and high-efficiency true random number generator is designed and implemented by using an analog-digital hybrid chaos model.A two-stage comparator is used instead of ADC to achieve chaos quantization and obtain true random bits to reduce resource consumption.A hyperchaotic system is constructed using a comb-shaped chaos map that is easy to implement in hardware as the basic unit.The simulation of chaotic system uses the optical chaos equation to achieve a simple and higher rate.Using FPGA and FPAA,the joint hardware implementation of true random number generator is carried out.The overall circuit is simple,and the true random sequence generated with relatively low resource cost has high stability and excellent statistical performance.It has passed the NIST SP800 standard randomness test and the most stringent TestU01 test.The design is expected to be integrated into special chip and widely used in various fields of network security.
关键词
真随机数发生器/模数混合/FPGA/超混沌Key words
true random number generator/analog digital hybrid/FPGA/hyper-chaos引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据