System support for robust data collection in wireless sensing systems

This dissertation studied how to provide system support for robust data collection in wireless sensing systems through addressing a few urgent design issues in the existing systems. A wireless sensing system may suffer issues arising at the sensors, during the data transmission, and during the data access by applications. Due to the unique characteristics of wireless sensing systems, certain conventional solutions for networked systems may not work well with these issues. We developed approaches to resolve these urgent problems in the design of wireless sensing systems. Specially, we have achieved the following: (1) we developed a resilient trust model to effectively detect faulty data in wireless sensing systems due to either sensor malfunctioning or malicious attempts to report false data; (2) we developed a low-cost, self-contained, accurate localization system for small-sized ground robotic vehicles, which enhances the wireless sensing systems containing mobile sensors by providing more accurate and highly available location data, with only limited overhead in economic cost and management; (3) we designed and implemented a robust trust-aware routing framework to secure multi-hop routing through a set of sensors in wireless sensing systems; (4) we developed a privacy-preserving wireless sensing system, which protects the user privacy while allowing arbitrary third-party applications to extract knowledge from the collected data