PUF Modeling Attacks on Simulated and Silicon Data

被引:406
作者
Ruehrmair, Ulrich [1 ]
Soelter, Jan [2 ]
Sehnke, Frank [3 ]
Xu, Xiaolin [4 ]
Mahmoud, Ahmed
Stoyanova, Vera [3 ]
Dror, Gideon [5 ]
Schmidhuber, Juergen [6 ,7 ]
Burleson, Wayne [4 ]
Devadas, Srinivas [8 ]
机构
[1] Tech Univ Munich, D-80333 Munich, Germany
[2] Free Univ Berlin, Inst Biol Neurobiol, D-14195 Berlin, Germany
[3] Tech Univ Munich, Dept Comp Sci, D-85748 Garching, Germany
[4] Univ Massachusetts, Dept Elect & Comp Engn, Amherst, MA 01003 USA
[5] Tel Aviv Yaffo Acad Coll, IL-61083 Tel Aviv Yaffo, Israel
[6] SUPSI Switzerland, CH-6928 Manno, Switzerland
[7] Univ Svizzera Italiana, CH-6900 Lugano, Switzerland
[8] MIT, Comp Sci & Artificial Intelligence Lab, Cambridge, MA 02139 USA
基金
美国国家科学基金会;
关键词
Physical unclonable functions; machine learning; cryptanalysis; physical cryptography; PHYSICAL UNCLONABLE FUNCTIONS; AUTHENTICATION; KEYS;
D O I
10.1109/TIFS.2013.2279798
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
We discuss numerical modeling attacks on several proposed strong physical unclonable functions (PUFs). Given a set of challenge-response pairs (CRPs) of a Strong PUF, the goal of our attacks is to construct a computer algorithm which behaves indistinguishably from the original PUF on almost all CRPs. If successful, this algorithm can subsequently impersonate the Strong PUF, and can be cloned and distributed arbitrarily. It breaks the security of any applications that rest on the Strong PUF's unpredictability and physical unclonability. Our method is less relevant for other PUF types such as Weak PUFs. The Strong PUFs that we could attack successfully include standard Arbiter PUFs of essentially arbitrary sizes, and XOR Arbiter PUFs, Lightweight Secure PUFs, and Feed-Forward Arbiter PUFs up to certain sizes and complexities. We also investigate the hardness of certain Ring Oscillator PUF architectures in typical Strong PUF applications. Our attacks are based upon various machine learning techniques, including a specially tailored variant of logistic regression and evolution strategies. Our results are mostly obtained on CRPs from numerical simulations that use established digital models of the respective PUFs. For a subset of the considered PUFs-namely standard Arbiter PUFs and XOR Arbiter PUFs-we also lead proofs of concept on silicon data from both FPGAs and ASICs. Over four million silicon CRPs are used in this process. The performance on silicon CRPs is very close to simulated CRPs, confirming a conjecture from earlier versions of this work. Our findings lead to new design requirements for secure electrical Strong PUFs, and will be useful to PUF designers and attackers alike.
引用
收藏
页码:1876 / 1891
页数:16
相关论文
共 51 条
  • [1] [Anonymous], 2007, Security, Privacy and Trust in Modern Data Management
  • [2] [Anonymous], J MACHINE LEARNING R
  • [3] [Anonymous], 2010, 14 INT WORK COMP EL, DOI DOI 10.1109/IWCE.2010.5677954
  • [4] [Anonymous], 2006, Pattern recognition and machine learning
  • [5] [Anonymous], 2001, Ph.D. thesis
  • [6] Back Thomas, 1996, EVOLUTIONARY ALGORIT
  • [7] LEARNABILITY AND THE VAPNIK-CHERVONENKIS DIMENSION
    BLUMER, A
    EHRENFEUCHT, A
    HAUSSLER, D
    WARMUTH, MK
    [J]. JOURNAL OF THE ACM, 1989, 36 (04) : 929 - 965
  • [8] Brzuska C., P CRYPTO 2011, P51
  • [9] Chen Q., P HOST 2011, P131
  • [10] Chen Q., P DATE 2012, P1459