Discrete Logarithmic Fuzzy Vault Scheme
This work addresses security vulnerabilities in biometric or cryptographic systems using fuzzy vaults, but it appears incremental as it builds on existing schemes with modifications.
The paper tackles the problem of enhancing security in fuzzy vault schemes by proposing three variants that integrate discrete logarithmic encryption, resulting in provable security improvements over classical fuzzy vaults.
In this paper a three fuzzy vault schemes which integrated with discrete logarithmic encryption scheme are proposed. In the first scheme, the message m is encoded with discrete logarithmic encryption scheme using randomly generated identity key \k{appa} for every message and then divided into non-overlapping segments. In the second scheme, the message is divided into non-overlapping segments and each segment is encoded with discrete logarithmic encryption scheme using the randomly generated identity key \k{appa}. In the third scheme, the message is divided into non-overlapping segments where even segments are encoded with identity key \k{appa}_even and odd segments are encoded with identity key \k{appa}_odd. Identity keys \k{appa}_even and \k{appa}_odd are randomly generated for every message. Finally, the encoded segments are declared as coefficients of a polynomial of specific degree. In all proposed schemes, elements of locking set A are used as X-coordinate values to compute evaluations of the polynomial by projecting elements of A onto points lying on the polynomial. A large number of random chaff points that do not lie on the polynomial are added to create noise to hide the encoded segments. Security analysis has shown the proposed scheme enjoys provable security over classical fuzzy vaults.