Security Defenses for Vulnerable Medical Sensor Network

Medical sensor networks have facilitated a wide range of applications in healthcare. However, these systems are in particular vulnerable to security attacks due to the fact that they are often not physically secured and are used in potentially hostile environments. We have proposed a theoretical and statistical framework for creating attacks and also the corresponding security defenses that include attack detection, diagnosis, and impact removal. We use medical shoes to collect data and demonstrate that low energy and low cost of medical sensor networks increase the probabilities of successful attacks. Our approach maps a semantic attack to an instance of an optimization problem where medical damage is maximized under the constraints of the probability of detection and root cause tracing which may consequence in incorrect medical diagnosis and treatment. Our results show that it is easy to attack several essential medical metrics and to alter corresponding medical diagnosis. Finally, we have developed several low energy and low overhead defense procedures for detecting and analyzing semantic security attacks.