Human-in-the-loop: Rethinking security in mobile and pervasive systems

Abstract

In this paper we argue that pervasive systems introduce human-driven security vulnerabilities that traditional usability design cannot address. We claim that there is a need to understand better the appropriate role of humans in the context of pervasive systems security, and to develop quantifiable and measurable concepts that describe humans and their relationship with our systems. Here, we highlight mobility and sociability as two new sources of security vulnerabilities for pervasive systems, and describe our method for developing quantifiable metrics for these concepts.