Wang, YinhaiFU, MINGJIAN2019-02-222019-02-222018FU_washington_0250O_19482.pdfhttp://hdl.handle.net/1773/43336Thesis (Master's)--University of Washington, 2018MAC address-based mobile device sensing is a popular data collection technology widely applied in transportation and other fields. This study focusses on improving existing MAC address-based mobile sensing data travel time measurement algorithm. Several major issues such as MAC address matching due to multiple detections, noises filtering due to different travel modes, and fluctuations due to low samples have been discussed and analyzed based on Acyclica data and License Plate Reader data in Seattle. A density-based clustering outlier filtering algorithm has been applied to improve travel time measurement accuracy. One of the upcoming challenges in MAC address-based travel time measurement due to MAC address randomization is analyzed, and the potential impact has been investigated by trip sampling rate and MAC collision rate.application/pdfen-USnoneMAC AddressMAC RandomizationMobile Device SensingTravel Time MeasurementTransportationComputer scienceCivil engineeringThesis