PPG-based biometric authentication: A review on architectures, datasets, attacks and security challenges.

Area of Science:

• Biometrics and Authentication
• Wearable Technology Security
• Signal Processing

Background:

• Photoplethysmography (PPG) signals are increasingly used for biometric authentication with wearable devices.
• Widespread adoption presents new security challenges and threats to identity verification systems.
• Existing research lacks a comprehensive analysis of PPG authentication security, particularly concerning attacks and dataset standardization.

Purpose of the Study:

• To present a taxonomy of PPG-based authentication and identification systems.
• To comparatively analyze authentication models based on PPG signals, considering performance, algorithms, hardware, and security.
• To focus on the impact and mitigation of attacks like spoofing and replay in PPG systems.

Main Methods:

• Review and taxonomy of existing PPG-based authentication systems.
• Comparative analysis of different authentication models and their security aspects.
• Identification and discussion of security gaps, including attack vectors and mitigation strategies.

Main Results:

• Identified a lack of standardized datasets, hindering reproducibility and generalization.
• Detailed the impact of spoofing, replay, and presentation attacks on PPG authentication.
• Highlighted the importance of standardized metrics (FAR, FRR, EER) and dataset standardization for scalability.

Conclusions:

• PPG-based authentication requires robust security measures and standardized evaluation protocols.
• Future work should focus on integrating advanced ML with standardized datasets and security evaluations for real-world reliability.
• Addressing dataset limitations and attack vectors is crucial for advancing secure PPG authentication.