Biometrie identification using photoplethysmography signal

Abstract

Due to the increase in the importance of the electronic data nowadays, finding a way to protect this data from hacking became a must. In this paper, the Photoplethysmographic signal known as the PPG signal is used as a biometric technique due to its several advantages, most importantly its unique form that differs between individuals. The PPG signal is taken using the Spo2 sensor which is a non-invasive method used for measuring the oxygen saturation from either the finger or the ear using 2 light emitting diodes (LEDs); using this PPG signal and its 2 derivatives 40 features depending on the signals' dimensions were extracted by Matlab, and then the K-nearest neighbor classifier was applied after tuning 2 parameters (the constant k and the distance metric) to check the efficiency of this method. The proposed algorithm was tested on dataset having signals previously processed. In future work, this technique is going to be implemented to unlock computers instead of using a password.