Journal
IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
Volume 8, Issue 11, Pages 1722-1730Publisher
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TIFS.2013.2277976
Keywords
Intellectual property protection; hardware metering; gate-level characterization
Funding
- NSF [CNS-0958369, CNS-1059435]
- Division Of Computer and Network Systems
- Direct For Computer & Info Scie & Enginr [1059435] Funding Source: National Science Foundation
Ask authors/readers for more resources
Hardware metering, the extraction of unique and persistent identifiers (IDs), is a crucial process for numerous integrated circuit (IC) intellectual property protection tasks. The currently known hardware metering approaches, however, are subject to alternations due to device aging, since they employ unstable manifestational IC properties. We, on the other hand, have developed the first robust hardware metering approach by using physical-level gate proprieties for ID generation. By using effective channel length, which is resilient to aging, and threshold voltage, which is essentially independent across gates and suitable for calculating the uniqueness of the IDs, we overcome the limitations of the existing approaches. Also, despite the increase in threshold voltage that occurs with aging, the original threshold voltage value can be extracted through intentional IC aging. Our ID generation procedure first employs two types of side channels, namely switching power and leakage power, to extract metering results for each gate. Next, we show that localized delay measurements alone are sufficient for accurate characterization of large sets of gates. Finally, by using threshold voltage for ID creation, we are able to obtain low probabilities of coincidence between legitimate and pirated ICs. The application of the approach to a set of benchmarks quantitatively establishes the effectiveness of the new hardware metering approach.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available