Reusable set constructions using randomized dissolvent templates for biometric security

  • Authors:
  • Jinyang Shi;Kwok-Yan Lam;Ming Gu;Husheng Li

  • Affiliations:
  • Key Laboratory for Information System Security, Ministry of Education, Tsinghua National Laboratory for Information Sci. & Tech., Tsinghua University, Beijing, P. R. China;Key Laboratory for Information System Security, Ministry of Education, Tsinghua National Laboratory for Information Sci. & Tech., Tsinghua University, Beijing, P. R. China;Key Laboratory for Information System Security, Ministry of Education, Tsinghua National Laboratory for Information Sci. & Tech., Tsinghua University, Beijing, P. R. China;Department of EECS, The University of Tennessee, Knoxville, TN

  • Venue:
  • GLOBECOM'09 Proceedings of the 28th IEEE conference on Global telecommunications
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

The emerging biometric cryptography has gained significant interests for key management and privacy protection, but the previously proposed schemes using set metrics for fingerprints may either be too weak to offer enough security or suffer from the performance limitations. In this paper, a new fuzzy cryptographic technique without use of chaff data, Randomized Dissolvent Template (RDT), is proposed for biometric set modalities. The proposed technique is designed to dissolve the enrolled biometric set into a random secret resource, so as to construct robust secured templates by exploiting at least two resources of randomness. In this way, when one fingerprint is used for multiple applications, each time the additional information leakage by secured templates will not exceed the new introduced random information, so RDT is reusable. We thus provide two novel RDT-based constructions in practice: Fuzzy Reconciler using set difference threshold and Fuzzy Dissolver using set intersection threshold. Security analysis proves the new constructions have enough computational complexity for the required security properties, and implementations on FVC2002DB fingerprint database show that the proposed schemes can bring about better accuracy performance over current Fuzzy Vault and Fuzzy Extractor, thus are more promising for biometric-based security applications.