Computer Science Colloquia
Thursday, April 10, 2014
Advisor: Gabriel Robins
Attending Faculty: John Stankovic, Chair; Worthy Martin, Yanjun Qi, and John Lach (Minor Representative)
9:30 AM, Rice Hall, Rm. 242
PhD Dissertation Defense Presentation
An RFID-Based Object Localization Framework and System
We explore a robust and scalable framework for locating objects using Radio Frequency Identification (RFID). Our object localization system remains accurate under a broad range of vendor hardware types, environmental conditions, and tag orientations. We demonstrate that radio signal strength, a parameter traditionally considered unreliable, can facilitate object localization. Our selection and binning techniques mitigate tag manufacturing variations and further improve localization precision. Extensive experiments indicate that our RFID-based approach can achieve deep sub-meter accuracy in localizing stationary as well as mobile objects in realistic environments.