Visible to the public A Novel Security Assessment Method Based on Linear Regression for Logic Locking

TitleA Novel Security Assessment Method Based on Linear Regression for Logic Locking
Publication TypeConference Paper
Year of Publication2019
AuthorsTsai, I-Chun, Zhong, Yi, Liu, Fang-Ru, Feng, Jianhua
Conference Name2019 IEEE International Conference on Electron Devices and Solid-State Circuits (EDSSC)
Keywordscomposability, cryptography, distribution probabilities, encrypted circuit, encrypted combinational benchmark circuits, encrypted key-inputs, hardware security, key-inputs modeling, linear regression, logic locking, logic locking security assessment method, Microelectronics Security, multiple linear regression, Predictive Metrics, probability, pubcrawl, Resiliency, security assessment
AbstractThis paper presents a novel logic locking security assessment method based on linear regression, by means of modeling between the distribution probabilities of key-inputs and observable outputs. The algorithm reveals a weakness of the encrypted circuit since the assessment can revoke the key-inputs within several iterations. The experiment result shows that the proposed assessment can be applied to varies of encrypted combinational benchmark circuits, which exceeds 85% of correctness after revoking the encrypted key-inputs.
DOI10.1109/EDSSC.2019.8754415
Citation Keytsai_novel_2019