Raw Data Raw Feature Data Discrete Feature Vector Reference Storage Behavioral or Physiological Trait Raw Data Raw Feature Data Discrete Feature Vector Data Acquisition Preprocessing Feature Extraction Store Feature Vector Authentication

Nowadays, biometric data are more and more used within authentication processes. These data are often stored in databases. However, these data underlie inherent privacy concerns. Therefore, special attention should be paid for handling of these data. We propose an extension of a similarity verification system with the help of the Paillier cryptosystem. In this paper, we use this system for signal processing in the encrypted domain for privacy-preserving biometric authentication. We adapt a biometric authentication system for enhancing privacy. We focus on performance issues with respect to database response time for our authentication process. Although encryption implicates computational effort, we show that only small computational overhead is required. Furthermore, we evaluate our implementation with respect to performance. However, the concept of verification of encrypted biometric data comes at the cost of increased computational effort in contrast to already available biometric systems. Nevertheless, currently available systems lack privacy enhancing technologies. Our findings emphasize that a focus on privacy in the context of user authentication is available. This solution leads to user-centric applications regarding authentication. As an additional benefit, results using data mining are more difficult to be obtained in the domain of user tracking.