Skip to main content

Behavior-based Mobility Research

Researchers: Patrick Tague, Joy Zhang

Research Area: Mobility

Abstract

CMU’s mobility research currently covers a variety of research topics ranging from wireless security, mobile-context awareness, airborne sensing, to in-building sensing, ultra-low-power computing, and indoor positioning systems. This research is motivated by the fact that increasingly more powerful user devices, such as mobile phones, in-vehicle and hand-held travel guidance systems are becoming ubiquitous mechanisms for Internet access and personalized computing. Wireless sensors embedded in appliances, vehicles, and physical environments are rapidly expanding Internet interactions with human users. However, context-aware services such as mobile shopping, advertising, gaming, and social networking could take advantage of wireless sensor networks only to the extent that the security and privacy of applications is preserved.

The utility of mobile networks can be enhanced by taking advantage of measurable or predictable user behavior; e.g., how people live, work, collaborate, and entertain themselves in mobile contexts. For example, CMU’s Silicon Valley campus has already created models of users’ behavior from observed mobile sensor time-series data using built-in accelerometers, gyroscopes, magnetometers, GPS subsystems, WiFi receivers, cameras, and microphones. Using these models, we can predict the future location of users, assist home-bound elderly individuals in real-time without relying on intrusive or costly care, and develop user authentication mechanisms that do not rely on secrets (e.g., passwords). However, we have not yet explored the security, privacy, and robustness implications of these models, and hence this constitutes a substantial unexplored area of fruitful future research.