A cyber-physical approach to trustworthy operation of health monitoring systems

摘要

Continuous health monitoring system (CHMS) are a collection of networked sensing devices that continuously monitor a user who is carrying them. The sensors can be worn by the user (e.g., fitbit or jawbone) or be part of a device that user carries (e.g., smartphones). Trustworthy operation is essential for CHMS due to the sensitive nature of the information they collect and the wireless transmission of the data to a sink/basestation entity for transport to a medical cloud for long term storage in a patient health record (PHR). In this regard, in the past we have proposed a scheme known as Physiological signal-based Key Agreement (PKA) to enable plug-n-play (i.e., transparent to the user in terms of configuration or setup) information security between wearable sensors that had access to same physiological signals (e.g., ECG, PPG). In this paper, we present Physiology-based System-wide Information Security (PySIS), which uses the concept of generative models (which generate synthetic physiological signals for a user) to extend PKA to enable end-to-end information security in CHMS from the sensors to the PHR. The crucial difference is that now we do not need to have access to the same physiological signals at both ends for our protocol to work. In addition, if PySIS fails and data leakage occurs in the system, we also propose a logging mechanism to perform forensic analysis of the system.