With the pervasiveness of mobile devices and the development of biometric technology, biometric identification, which can achieve individual authentication relies on personal biological or behavioral characteristics, has attracted widely considerable interest. However, privacy issues of biometric data bring out increasing concerns due to the highly sensitivity of biometric data. Aiming at this challenge, in this paper, we present a novel privacy-preserving online fingerprint authentication scheme, named e-Finga, over encrypted outsourced data. In the proposed e-Finga scheme, the user's fingerprint registered in trust authority can be outsourced to different servers with user's authorization, and secure, accurate and efficient authentication service can be provided without the leakage of fingerprint information. Specifically, an improved homomorphic encryption technology is introduced to achieve an efficient online fingerprint matching algorithm over encrypted FingerCode data in the outsourcing scenarios. Through detailed security analysis, we show that e-Finga can resist various security threats. In addition, we implement e-Finga over a workstation with a real fingerprint database, and extensive simulation results demonstrate that the proposed e-Finga scheme can serve efficient and accurate online fingerprint authentication.
FingerCode-based id matching
System diagram of the FingerCode-based id matching
(a) Encryption Template Authorization
(b) Fingerprint Matching
Computation complexity and integrated performance
(a) Computation cost of OASer with different FingerCode dimensions.
(b) Computation cost of TA with different FingerCode dimensions.
(c) Computation cost of user with different FingerCode dimensions.
(d) Query and response time with different FingerCode dimensions.
Query and response time in e-Finga and PFRS