An improved structure of fuzzy extractor without entropy leakage

Abstract

The physical unclonable function(PUF) module cannot be directly applied to key generation due to the instability of the response, and the processing method of the fuzzy extractor brings the security risk of entropy leakage due to the disclosure of the helper data. An improved fuzzy extractor structure is proposed for the entropy leakage problem of fuzzy extractor, by adding artificial noise channel, the modified PUF response is unbiased, and the effect of non-entropy leakage is realized without increasing the resource-constrained device operations in the reproduction phase. In the context of SRAM PUF with bias pb=0.54 and error rate e=0.10, the required SRAM PUF size and key generation success rate are improved compared to other methods.

Keywords

• physical unclonable function
• fuzzy extractor
• entropy leak
• key generation