Secure Biometric Authentication Systems: Trends and Challenges on Privacy and Security

Abstract

The use of biometric authentication systems in the contemporary security applications has been adopted by many because of the reliability and ease of usage that they offer when it comes to identity verification. However, the growing use of biometric technologies in the large-scale and networked environments has raised the acute issues associated with the security vulnerabilities and privacy in the user domain. The literature review research is mostly conducted in terms of modality-based comparisons or performance assessment, whereas the connection between security threats and privacy-preserving mechanisms is not thoroughly studied. The paper provides an extensive threat-based review of the biometric authentication systems with the focus on privacy maintenance. In addition, the paper also discusses the latest privacy-saving methods, such as cancellable biometrics, biometric cryptosystems, homomorphic encryption, secure multi-party computation, and privacy-conscious machine learning methods. The effectiveness and limitation of the existing approaches is highlighted by the use of a structured threat-to-solution mapping.