Privacy-Preserving Comparison of Variable-Length Data With Application to Biometric Template Protection
The establishment of cloud computing and big data in a wide variety of daily applications has raised some privacy concerns due to the sensitive nature of some of the processed data. This has promoted the need to develop data protection techniques, where the storage and all operations are carried out without disclosing any information. Following this trend, this paper presents a new approach to efficiently compare variable-length data in the encrypted domain using homomorphic encryption where only encrypted data is stored or exchanged. The new variable-length-based algorithm is fused with existing fixed-length techniques in order to obtain increased comparison accuracy. To assess the soundness of the proposed approach, we evaluate its performance on a particular application: a multi-algorithm biometric template protection system based on dynamic signatures that complies with the requirements described in the ISO/IEC 24745 standard on biometric information protection. Experiments have been carried out on a publicly available database and a free implementation of the Paillier cryptosystem to ensure reproducibility and comparability to other schemes.
Privacy, security, homomorphic encryption, template protection, biometrics, signature.