As wireless networks connect computers that are embedded in our cars, cell phones, and personal items, a key challenge is the preservation of users' location privacy while enabling provisioning of useful services. Wireless networks' shared medium and unique device identifiers implicitly enable location tracking. Even if the mapping to individuals' identity is initially unknown, the accumulation of location traces will eventually allow identifying users -- for example through correlation with home or work address databases. Especially the time-series nature of such location traces poses novel privacy challenges that have not been addressed by more general data privacy mechanisms. This project develops multi-layer anonymity techniques, trajectory-aware identifier switching at the network link layer and time-series data perturbation techniques at the application layer. The project experimentally evaluates these techniques' resilience against advanced localization and target tracking mechanisms using a large-scale wireless network research testbed.
The project's results are expected to inform emerging privacy standards and policy decisions, for example at the FCC. The project may uncover powerful attacks on user privacy that warrant stronger legal deterrence; or the project may develop efficient protection mechanisms that could be mandated for service providers. Furthermore, engagement in privacy research will encourage students to reflect on the ethics of the computer engineering profession.
