This project addresses the challenge of providing usable (plug-and-play, self-configuring, and autonomic) security solutions for Body Area Networks (BANs): networks of economically powered, wireless, wearable and/or implanted health monitoring nodes (sensors and actuators), for collecting and communicating health information and, appropriately administering medicine or prosthetic actions. BANs have many diverse applications including sports health management, home-based health-care and post-operative care. Due to the sensitive nature of the data collected, securing BANs is important for privacy preservation and protecting the host from bodily harm. Traditional security methods require considerable setup efforts in terms of pre-deployment of cryptographic secrets (key distribution). In contrast, this project takes a cyber-physical approach and uses the concept of Physiological Value based Security (PVS), which employs physiological values (PVs) to generate cryptographic keys for securing BANs. Two important issues in this project are: 1) identification of appropriate PVs for generating high-quality keys; and 2) development of signal processing techniques to generate cryptographic keys from the physiological values. For the former issue, the use of common PVs (e.g. Electrocardiogram, Photoplethysmogram, etc.) would be investigated; for the latter, time and frequency domain feature generation techniques would be developed. This project will result in: a) benchmarks for selecting appropriate PVs, b) light-weight measurement and synchronization algorithms to extract PVs in real-time, c) a software and hardware (FPGA-based) prototype implementation of PVS in IMPACT lab?s AYUSHMAN pervasive health monitoring testbed. These results (to be made publicly available at http://impact.asu.edu) will be an important step toward providing dependable, affordable healthcare for all.
