1. 长沙理工大学物理与电子科学学院,湖南,长沙,410114
2. 柔性电子材料基因工程湖南省重点实验室, 湖南长沙 4100114
3. 长沙理工大学物理与电子科学学院,湖南,长沙,410114
4. 柔性电子材料基因工程湖南省重点实验室 湖南长沙 4100114
网络出版:2019-10-25,
纸质出版:2019
移动端阅览
白创, 唐立军. 一种可靠的芯片指纹PUF电路[J]. 电子学报, 2019,47(10):2116-2125.
BAI Chuang, TANG Li-jun. A Reliable Physical Unclonable Function for Chip Fingerprint[J]. Acta Electronica Sinica, 2019, 47(10): 2116-2125.
白创, 唐立军. 一种可靠的芯片指纹PUF电路[J]. 电子学报, 2019,47(10):2116-2125. DOI: 10.3969/j.issn.0372-2112.2019.10.013.
BAI Chuang, TANG Li-jun. A Reliable Physical Unclonable Function for Chip Fingerprint[J]. Acta Electronica Sinica, 2019, 47(10): 2116-2125. DOI: 10.3969/j.issn.0372-2112.2019.10.013.
本文引入一种可靠的芯片指纹物理不可克隆函数(Physical Unclonable Function)电路.该PUF包括基于电流饥饿型延迟单元的工艺敏感电路、时间偏差放大器、时间偏差比较器、表决机制与扩散算法五个部分.通过捕获制造工艺的偏差,每一个工艺敏感电路可以稳定产生两路具有微弱延时差的延迟信号,然后比较生成指纹ID;设计一种新型的扩散算法改善PUF的唯一性,引入时间偏差放大器与表决机制增强PUF相对于温度与电源电压变化的稳定性.文中PUF在0.18m CMOS工艺下实现.仿真结果表明,该PUF的输出具有均匀统计分布特征,同时在温度从-40℃至100℃,电源电压从1.7V至1.9V变化条件下,其输出ID具有97.5%的稳定性.
A reliable physical unclonable function (PUF) for chip fingerprint using current starved delay element (CSDE) is described in this paper. The proposed PUF is composed of CSDE-based sensors
time difference amplifier
time difference comparator
voting mechanism and diffusion algorithm circuit. By capturing manufacturing process variations
each sensor produces two slightly different delay-time values that can be compared in order to create a digital identification (ID) for the chip. A new diffusion algorithm is designed to further improve the uniqueness of PUFs. Time difference amplifier and voting mechanism are introduced to simultaneously improve the reliability of PUFs across temperature and supply voltage variations. The proposed PUF is designed in 0.18μm CMOS technology. Simulated results show that the proposed PUF has a good output statistical characteristic of uniform distribution
and a high stability of 97.5% with respect to temperature variation from -40℃ to 100℃
and supply voltage variation from 1.7V to 1.9V.
0
浏览量
117
下载量
1
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621