Biometric Authentication Using Smartphone-Generated Acoustic Signals Modulated by Vascular Dynamics

Smartphones now play a central role in high-value transactions and IoT control, making reliable, low-cost, and privacy-preserving authentication essential. Existing vision- and PPG-based biometrics are sensitive to ambient conditions or burdensome to use, highlighting the need for alternatives with built-in liveness detection. We present EchoAuth, a vascular biometric system that leverages only a smartphone's speaker and microphone to sense wrist hemodynamics. By actively probing the wrist and applying wavelet-based denoising to reduce motion artifacts while preserving pulse morphology, EchoAuth produces high-fidelity signals for feature extraction and classification. It achieves satisfactory performance with inherent liveness sensing from physiological pulsation, demonstrating the potential for secure, unobtrusive, and universally deployable mobile authentication.