Feature-level fusion of physiological parameters to be used as cryptographic keys

Abstract

In this paper, we propose two novel feature-level fused physiological parameter generation techniques: (i) concat-fused physiological parameter generation, and (ii) xor-fused physiological parameter generation, output of which can be used to secure the communication among the biosensors in Body Area Network (BAN). In these physiological parameter generation techniques, we combine a time-domain physiological parameter with a frequency-domain physiological parameter, in order to achieve robust performance compared to their singular versions. We analyze both the performance and the quality of the outcomes. Our results show that we generate good candidates of physiological parameters that can be used as cryptographic keys to provide security for the intra-network communication in BANs.