Large-Scale Tracking of Smartphone Users

Abstract

In a world where technology is becoming increasingly important and widespread, information security, especially that which concerns personal devices and smartphones is critical in protecting users from malicious attacks and privacy breaches. A group of researchers at Princeton seeking to find such breaches, wrote an application, PinMe, that could, by analyzing seemingly innocuous data from a smartphone collected without a user’s permission, somewhat accurately determine the location of a user without utilizing location services. This study, however, was done with only a few test users. My research had two goals: To determine whether this tracking could be done in realtime with many users, and whether this tracking method could be used for navigational services without the use of a GPS. To that end, I designed an Android application to test PinMe’s algorithm in real time and help a user navigate in the real world. Through testing, I concluded that, due to background restrictions imposed by Android 6.0+, achieving the required sampling rate of sensors for tracking users is not possible in a practical setting and thus not practically feasible on a large scale. Likewise, an increase in these restrictions starting from Android 8.0 inhibits the ability of the application from gathering data for navigation. This, as well as the limitations of the algorithm promise a lousy user experience in the navigation portion of the app, though, under some strict assumptions, navigation is likely possible.