信息网络安全 ›› 2022, Vol. 22 ›› Issue (8): 36-43.doi: 10.3969/j.issn.1671-1122.2022.08.005

• 技术研究 • 上一篇    下一篇

抗温度干扰的STT-MRAM随机数生成器及其安全性分析

伍麟珺1, 刘洋2, 袁涛3, 胡玉鹏1()   

  1. 1.湖南大学信息科学与工程学院,长沙 410012
    2.湖南省交通运输厅科技信息中心,长沙 410004
    3.湖南国科微电子股份有限公司,长沙 410100
  • 收稿日期:2021-11-29 出版日期:2022-08-10 发布日期:2022-09-15
  • 通讯作者: 胡玉鹏 E-mail:yphu@hnu.edu.cn
  • 作者简介:伍麟珺(1982—),女,湖南,博士研究生,主要研究方向为硬件安全、人工智能安全与隐私保护|刘洋(1973—),男,湖南,工程师,硕士,主要研究方向为信息安全|袁涛(1976—),男,湖南,高级工程师,硕士,主要研究方向为大规模集成电路设计及信息安全|胡玉鹏(1981—),男,湖南,教授,博士,主要研究方向为大数据、人工智能、云计算等。
  • 基金资助:
    国家自然科学基金(61872130);国家自然科学基金联合基金重点项目(U20A20202);国家自然科学基金面上项目(61874042)

STT-MRAM Random Number Generator with Anti-Temperature Interference and Security Analysis

WU Linjun1, LIU Yang2, YUAN Tao3, HU Yupeng1()   

  1. 1. College of Computer Science and Electronic Engineering, Hunan University, Changsha 410012, China
    2. Science and Technology Information Center of Hunan Transportation Department, Changsha 410004, China
    3. Hunan Guoke Microelectronics Co., Ltd., Changsha 410100, China
  • Received:2021-11-29 Online:2022-08-10 Published:2022-09-15
  • Contact: HU Yupeng E-mail:yphu@hnu.edu.cn

摘要:

近年来,各向异性磁性材料因良好的随机特性为随机数发生器(Random Number Generator,RNG)等重要的硬件安全原语设计提供了一种新思路。已有的基于磁隧道结(Magnetic Tunnel Junction,MTJ)的随机数发生器方案虽然具有更高的安全性、能效和集成度等优点,但依然无法有效解决输出序列随机性受温度影响的问题。文章提出了非均匀写入法和非固定参考法两种灵活的抗温度干扰的真随机数产生方法。两种方法在提升随机数电路输出随机性的同时尽可能抵消环境温度的干扰。实验结果表明,两种随机数产生方案产生的随机数的香农熵在97%左右,且以较高的通过率(>98.5%)通过美国国家标准与技术研究院(National Institute of Standards and Technology,NIST)测试。

关键词: 随机数生成器, 磁随机存储器, 抗温度干扰, 非均匀写入, 非固定参考

Abstract:

In recent years, anisotropic magnetic materials are being used to build essential security primitives like random number generators (RNG). The existing magnetic tunnel junction (MTJ)-based random number generator scheme, although with higher security, energy efficiency, and integration density, cannot solve the problem effectively that the randomness of the output is affected by temperature. This paper presents two novel random number generator methods against temperature interference, non-uniform writing and uncertain reference. The two methods improve the randomness of random number outputs while suppressing the interference caused by environmental temperature. The experimental results demonstrate that the Shannon entropy of the random numbers generated by the two RNG schemes is around 97%. The proposed RNG can pass the National Institute of Standards and Technology (NIST) test suites with high passing rates (>98.5%).

Key words: random number generator, STT-MRAM, anti-temperature interference, non-uniform writing, uncertain reference

中图分类号: