Peak misdetection in heart-beat-based security

Characterization and tolerance

Conference Paper (2014)
Author(s)

Robert M. Seepers (Erasmus MC)

Christos Strydis (Erasmus MC)

Pedro Peris-Lopez (Carlos III University of Madrid)

Ioannis Sourdis (Chalmers University of Technology)

Chris I. De Zeeuw (Erasmus MC)

Affiliation
External organisation
DOI related publication
https://doi.org/10.1109/EMBC.2014.6944847 Final published version
More Info
expand_more
Publication Year
2014
Language
English
Affiliation
External organisation
Article number
6944847
Pages (from-to)
5401-5405
ISBN (electronic)
9781424479290
Event
2014 36th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2014 (2014-08-26 - 2014-08-30), Chicago, United States
Downloads counter
215

Abstract

The Inter-Pulse-Interval (IPI) of heart beats has previously been suggested for security in mobile health (mHealth) applications. In IPI-based security, secure communication is facilitated through a security key derived from the time difference between heart beats. However, there currently exists no work which considers the effect on security of imperfect heart-beat (peak) detection. This is a crucial aspect of IPI-based security and likely to happen in a real system. In this paper, we evaluate the effects of peak misdetection on the security performance of IPI-based security. It is shown that even with a high peak detection rate between 99.9% and 99.0%, a significant drop in security performance may be observed (between -70% and -303%) compared to having perfect peak detection. We show that authenticating using smaller keys yields both stronger keys as well as potentially faster authentication in case of imperfect heart beat detection. Finally, we present an algorithm which tolerates the effect of a single misdetected peak and increases the security performance by up to 155%.