Experimentation-Based Sampling Scheme for GPS-Smartphone Probe Vehicles

Abstract

Global positioning system (GPS)-smartphones as probe vehicles have become a promising technique for real-time traffic surveillance and dynamic traffic information collection. In order to achieve cost-effective traffic information collection and estimation, a suitable sampling interval should be determined. This paper investigates two optimal sampling intervals, including space-based and time-based sampling intervals. With the requirements of high-link travel speed estimation accuracy and map-matching accuracy, the paper proposes optimal space-based and time-based sampling interval (around 120m and 20s) for GPS-smartphone probes.