Photoplethysmography Data Reduction Using Truncated Singular Value Decomposition and Internet of Things Computing

Biometric-based identity authentication is integral to modern-day technologies. From smart phones, personal computers, and tablets security checkpoints, they all utilize a form of check based on methods such as face recognition fingerprint-verification. Photoplethysmography (PPG) another biometric-based that has recently been gaining momentum, because it effective easy implement. This paper considers cloud-based system model for PPG-authentication, where the PPG signals various individuals are collected with distributed sensors communicated cloud authentication. Such incursarge signal traffic, especially in crowded places airport checkpoints. motivates need compression–decompression scheme (or Codec short). The required reduce data traffic by compressing each before communicated, i.e., encoding right after comes off sensor sent be reconstructed (i.e., decoded). Therefore, two requirements meet: (i) produce high-fidelity reconstruction; (ii) have computationallyightweight encoder. Both met proposed this paper, which designed using truncated singular value decomposition (T-SVD). developed tested publicly available dataset from multiple individuals, namely CapnoBase dataset. It shown achieve 95% compression ratio 99% coefficient determination. means capable delivering first requirement, reconstruction, while producing highly compressed signals. Those do not require heavy computations produced well. An implementation single-board computer attempted encoder, showing encoder can average 300 milliseconds per Raspberry Pi 3. enough time encode prior transmission cloud.