Lead Institution: Dartmouth

Project Leader: David Kotz

Research Progress

• Abstract
  The purpose of this project was to determine whether two devices are physically on the same body. We developed an algorithm and a successful prototype to accomplish this goal. Our approach assumes each device has a built-in motion sensor (accelerometer and optional gyroscope) and a short-range wireless radio. We show our method is capable of detecting whether two devices are attached to the same body, during any period when the person wearing the two devices is walking.

• Focus of the research/Market need for this project
  There are no commercial products that offer this functionality. We imagine our method could be integrated into many wearable devices.

• Project Aims/Goals
  We investigated methods for determining whether two (or more) sensors are on the same body of an individual. For example, a smartphone acting as a sensor ought to be able to determine if another sensor within communication range is attached to the same human body. This method is useful in concert with our research on recognizing the wearer of a sensor because it allows us to create a specialized device (e.g., an Amulet) that can take care of providing all the properties we desire (e.g., provenance, security, privacy) while allowing other sensors to perform their specialized sensing function. Our method requires minimal overhead: the inclusion of an accelerometer in the sensing device. The smartphone can then use its own accelerometer and our algorithms to determine if the sensor is on the same body of the smartphone.

• Key Conclusions/Significant Findings/Milestones reached/Deliverables
  We published our methods, and a rigorous evaluation, in a conference paper, a journal paper, and a PhD dissertation. Our journal publication has a rigorous analysis of our method and a small study showing how often our method could be run over the course of the day given that we require subjects to be walking. In Cornelius’ dissertation, he extended the method to include a gyroscope sensor under the belief that it would add additional performance. He found that the naively combined accelerometer and gyroscope data streams do not perform better than just an accelerometer alone.

• Materials Available for Other Investigators/interested parties
  Peer-reviewed published papers.

• Market entry strategies
  None.

Bibliography
Usable Security for Wireless Body-Area Networks
Cory Cornelius; PhD dissertation
Dartmouth Computer Science Technical Report TR2013-741, September 2013

Recognizing Whether Sensors are on the Same Body
Cory Cornelius and David Kotz
Journal of Pervasive and Mobile Computing (PMC), 8(6); 822-836, December 2012

Recognizing Whether Sensors are on the Same Body
Cory Cornelius and David Kotz
Proceedings of the International Conference on Pervasive Computing, volume 6696 of Lecture Notes in Computer Science, pages 332–349. Springer-Verlag, June 2011