Smart wireless devices are increasingly becoming the preferred method for accessing physical locations, bank accounts, and other valuable resources. However, their use for authentications comes with increased security considerations, since they can be stolen and are vulnerable to different types of software attacks. The objective of the research is to develop a new type of pervasive authentication for smart wireless networked systems based on macroscopic human behavior. The key hypothesis is that the regularity in a person's behavior can be used to uniquely identify her, as well as to substitute static passwords with one-time authentication challenges pertaining to her past behavior. To this end, this project will focus in three integrated components: (1) methodologies for behavioral data collection and authentication, spanning from physical layer fingerprinting techniques to privacy-preserving authentication protocols; (2) algorithms for discovering and modeling the regularity in a user's behavior to pervasively authenticate her; and (3) methods for generating one-time authentication challenges based on the user's preferences, schedules and routines. The expected result of the research is a radically new paradigm in authentication that will eliminate the risks associated with smart wireless devices, and can also find application in other domains such as online authentication. The research outcomes and parts of the developed authentication framework will be disseminated to the widest audience possible, including application developers, through the project?s website. The educational component of the project includes integration of the research agenda with curriculum development, and outreach activities that engage high-school students and underrepresented communities in computer science and engineering.
