Machine learning attacks on 65nm Arbiter PUFs: Accurate modeling poses strict bounds on usability. Hospodar, G., Maes, R., & Verbauwhede, I. In 2012 IEEE International Workshop on Information Forensics and Security (WIFS), pages 37–42, December, 2012. doi abstract bibtex Arbiter Physically Unclonable Functions (PUFs) have been proposed as efficient hardware security primitives for generating device-unique authentication responses and cryptographic keys. However, the assumed possibility of modeling their underlying challenge-response behavior causes uncertainty about their actual applicability. In this work, we apply well-known machine learning techniques on challenge-response pairs (CRPs) from 64-stage Arbiter PUFs realized in 65nm CMOS, in order to evaluate the effectiveness of such modeling attacks on a modern silicon implementation. We show that a 90%-accurate model can be built from a training set of merely 500 CRPs, and that 5000 CRPs are sufficient to perfectly model the PUFs. To study the implications of these attacks, there is need for a new methodology to assess the security of PUFs suffering from modeling. We propose such a methodology and apply it to our machine learning results, yielding strict bounds on the usability of Arbiter PUFs. We conclude that plain 64-stage Arbiter PUFs are not secure for challenge-response authentication, and the number of extractable secret key bits is limited to at most 600.
@inproceedings{hospodar_machine_2012,
title = {Machine learning attacks on 65nm {Arbiter} {PUFs}: {Accurate} modeling poses strict bounds on usability},
shorttitle = {Machine learning attacks on 65nm {Arbiter} {PUFs}},
doi = {10.1109/WIFS.2012.6412622},
abstract = {Arbiter Physically Unclonable Functions (PUFs) have been proposed as efficient hardware security primitives for generating device-unique authentication responses and cryptographic keys. However, the assumed possibility of modeling their underlying challenge-response behavior causes uncertainty about their actual applicability. In this work, we apply well-known machine learning techniques on challenge-response pairs (CRPs) from 64-stage Arbiter PUFs realized in 65nm CMOS, in order to evaluate the effectiveness of such modeling attacks on a modern silicon implementation. We show that a 90\%-accurate model can be built from a training set of merely 500 CRPs, and that 5000 CRPs are sufficient to perfectly model the PUFs. To study the implications of these attacks, there is need for a new methodology to assess the security of PUFs suffering from modeling. We propose such a methodology and apply it to our machine learning results, yielding strict bounds on the usability of Arbiter PUFs. We conclude that plain 64-stage Arbiter PUFs are not secure for challenge-response authentication, and the number of extractable secret key bits is limited to at most 600.},
booktitle = {2012 {IEEE} {International} {Workshop} on {Information} {Forensics} and {Security} ({WIFS})},
author = {Hospodar, G. and Maes, R. and Verbauwhede, I.},
month = dec,
year = {2012},
pages = {37--42}
}
Downloads: 0
{"_id":"hToR6BuFYkDR9H5xB","bibbaseid":"hospodar-maes-verbauwhede-machinelearningattackson65nmarbiterpufsaccuratemodelingposesstrictboundsonusability-2012","authorIDs":[],"author_short":["Hospodar, G.","Maes, R.","Verbauwhede, I."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Machine learning attacks on 65nm Arbiter PUFs: Accurate modeling poses strict bounds on usability","shorttitle":"Machine learning attacks on 65nm Arbiter PUFs","doi":"10.1109/WIFS.2012.6412622","abstract":"Arbiter Physically Unclonable Functions (PUFs) have been proposed as efficient hardware security primitives for generating device-unique authentication responses and cryptographic keys. However, the assumed possibility of modeling their underlying challenge-response behavior causes uncertainty about their actual applicability. In this work, we apply well-known machine learning techniques on challenge-response pairs (CRPs) from 64-stage Arbiter PUFs realized in 65nm CMOS, in order to evaluate the effectiveness of such modeling attacks on a modern silicon implementation. We show that a 90%-accurate model can be built from a training set of merely 500 CRPs, and that 5000 CRPs are sufficient to perfectly model the PUFs. To study the implications of these attacks, there is need for a new methodology to assess the security of PUFs suffering from modeling. We propose such a methodology and apply it to our machine learning results, yielding strict bounds on the usability of Arbiter PUFs. We conclude that plain 64-stage Arbiter PUFs are not secure for challenge-response authentication, and the number of extractable secret key bits is limited to at most 600.","booktitle":"2012 IEEE International Workshop on Information Forensics and Security (WIFS)","author":[{"propositions":[],"lastnames":["Hospodar"],"firstnames":["G."],"suffixes":[]},{"propositions":[],"lastnames":["Maes"],"firstnames":["R."],"suffixes":[]},{"propositions":[],"lastnames":["Verbauwhede"],"firstnames":["I."],"suffixes":[]}],"month":"December","year":"2012","pages":"37–42","bibtex":"@inproceedings{hospodar_machine_2012,\n\ttitle = {Machine learning attacks on 65nm {Arbiter} {PUFs}: {Accurate} modeling poses strict bounds on usability},\n\tshorttitle = {Machine learning attacks on 65nm {Arbiter} {PUFs}},\n\tdoi = {10.1109/WIFS.2012.6412622},\n\tabstract = {Arbiter Physically Unclonable Functions (PUFs) have been proposed as efficient hardware security primitives for generating device-unique authentication responses and cryptographic keys. However, the assumed possibility of modeling their underlying challenge-response behavior causes uncertainty about their actual applicability. In this work, we apply well-known machine learning techniques on challenge-response pairs (CRPs) from 64-stage Arbiter PUFs realized in 65nm CMOS, in order to evaluate the effectiveness of such modeling attacks on a modern silicon implementation. We show that a 90\\%-accurate model can be built from a training set of merely 500 CRPs, and that 5000 CRPs are sufficient to perfectly model the PUFs. To study the implications of these attacks, there is need for a new methodology to assess the security of PUFs suffering from modeling. We propose such a methodology and apply it to our machine learning results, yielding strict bounds on the usability of Arbiter PUFs. We conclude that plain 64-stage Arbiter PUFs are not secure for challenge-response authentication, and the number of extractable secret key bits is limited to at most 600.},\n\tbooktitle = {2012 {IEEE} {International} {Workshop} on {Information} {Forensics} and {Security} ({WIFS})},\n\tauthor = {Hospodar, G. and Maes, R. and Verbauwhede, I.},\n\tmonth = dec,\n\tyear = {2012},\n\tpages = {37--42}\n}\n\n","author_short":["Hospodar, G.","Maes, R.","Verbauwhede, I."],"key":"hospodar_machine_2012","id":"hospodar_machine_2012","bibbaseid":"hospodar-maes-verbauwhede-machinelearningattackson65nmarbiterpufsaccuratemodelingposesstrictboundsonusability-2012","role":"author","urls":{},"downloads":0},"bibtype":"inproceedings","biburl":"https://bibbase.org/zotero/ky25","creationDate":"2019-05-11T17:47:04.701Z","downloads":0,"keywords":[],"search_terms":["machine","learning","attacks","65nm","arbiter","pufs","accurate","modeling","poses","strict","bounds","usability","hospodar","maes","verbauwhede"],"title":"Machine learning attacks on 65nm Arbiter PUFs: Accurate modeling poses strict bounds on usability","year":2012,"dataSources":["XxiQtwZYfozhQmvGR"]}