• DocumentCode
    254363
  • Title

    When a Bloom filter is a Doom filter: Security assessment of a novel iris biometric template protection system

  • Author

    Hermans, J. ; Mennink, B. ; Peeters, R.

  • fYear
    2014
  • fDate
    10-12 Sept. 2014
  • Firstpage
    1
  • Lastpage
    6
  • Abstract
    Biometric template protection systems are expected to meet two major security requirements: irreversibility and unlinkability. We analyze the Bloom filter based iris biometric template protection system recently introduced by Rathgeb et al. at ICB 2013 and IET Biometrics 2014. We demonstrate that the scheme does not achieve unlinkability, presenting a simple attack that in the worst case succeeds with probability at least 96%. We also present a security analysis on generating false positives or recovering the key, both leading to undesirably low attack complexities: 225 for generating false positives for the smaller versions of the scheme, and a complexity between 22 and 28 for recovering the secret key.
  • Keywords
    data structures; iris recognition; probability; ICB 2013; IET Biometrics 2014; bloom filter; doom filter; low attack complexities; novel iris biometric template protection system; probability; secret key; security analysis; security assessment; Complexity theory; Feature extraction; Hamming weight; Indexes; Iris recognition; Security; Vectors;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Biometrics Special Interest Group (BIOSIG), 2014 International Conference of the
  • Conference_Location
    Darmstadt
  • Print_ISBN
    978-3-88579-624-4
  • Type

    conf

  • Filename
    7029413