Unlock Your Heart: Next Generation Biometric in Resource-Constrained Healthcare Systems and IoT

With the emergence of the Internet-of-Things, there is a growing need for access control and data protection on low-power, pervasive devices. Key-based biometric cryptosystems are promising for IoT due to its convenient nature and lower susceptibility to attacks. However, the costs associated with biometric processing and template protection are nontrivial for smart cards, and so forth. In this paper, we discuss the cost versus the utility of biometric systems and investigate frameworks for improving them. We propose the noise-aware biometric quantization framework (NA-IOMBA) capable of generating unique, reliable, and high entropy keys with low enrollment times and costs with several experiments. First, we compare its performance with IOMBA and one-class-SVM on multiple biometric modalities, including popular ones (fingerprint and iris) and emerging cardiovascular ones (ECG and PPG). The results show that NA-IOMBA outperforms them all and that ECG provides the best trade-off between reliability, key length, entropy, and implementation cost. Second, we examine the impact on key reliability with ECGs obtained at different sessions and trained with a different number of heartbeats. Finally, implementation results show that incorporating noise models with NA-IOMBA reduces power and utilization overhead by more than 60% by adapting the pre-processing, feature extraction, and post-processing modules.