A Post-Quantum Fuzzy Commitment Scheme for Biometric Template Protection: An Experimental Study

The proliferation and extensive acceptance of biometrics in many sectors is due to its merits over traditional password-based authentication. Several biometric template protection schemes have been proposed during the last decade to address the security and privacy issues related to biometrics. While error-correcting coding is required in the design of these schemes, it has been shown to degrade security and accuracy performance. In this study, a new post-quantum fuzzy commitment scheme (PQFC) for biometric template protection is designed, which is proved to be secure based on the Short Vector Problem (SVP) of the lattice. The present work provides the first systematic construction of biometric template protection that does not rely on error-correcting codes. To that purpose, a novel biometric-based PQFC authentication system based on a smartcard is created, with the biometric template protected by PQFC. Furthermore, biometric matching is done indirectly. The biometric-based PQFC authentication system is examined using an iris database of 108 individuals. According to the experimental data, the accuracy of iris recognition with PQFC is 99.1 percent, which is higher than the accuracy without PQFC, which is 97.4 percent. A False Accepting Rate (FAR) of 0% is achieved, while the False Rejection Rate (FRR) is 2.9%. The FRR can be adjusted by employing the best standard iris recognition algorithm. The execution time for the essential operations of PQFC and the storage requirements are estimated. Furthermore, the security of the proposed authentication system is analyzed, showing that the renewability, unlinkability, and irreversibility of the ISO/IEC 24745 standard are fulfilled without downgrading the accuracy. The overall performance of this work demonstrated that the PQFC is a promising technique for biometric template protection that is also suitable for the Internet of Things and cloud computing applications.