EKG-based key agreement in Body Sensor Networks

Preserving a person's privacy in an efficient manner is very important for critical, life-saving infrastructures like body sensor networks (BSN). This paper presents a novel key agreement scheme which allows two sensors in a BSN to agree to a common key generated using electrocardiogram (EKG) signals. This EKG-based key agreement (EKA) scheme aims to bring the "plug-n-play" paradigm to BSN security whereby simply deploying sensors on the subject can enable secure communication, without requiring any form of initialization such as pre-deployment. Analysis of the scheme based on real EKG data (obtained from MIT PhysioBank database) shows that keys resulting from EKA are: random, time variant, can be generated based on short-duration EKG measurements, identical for a given subject and different for separate individuals.