Accurate, low-energy trajectory mapping for mobile devices

摘要: CTrack is an energy-efficient system for trajectory mapping using raw position tracks obtained largely from cellular base station fingerprints. Trajectory mapping, which involves taking a sequence of samples and producing the most likely path followed by user, important component in many location-based services including crowd-sourced traffic monitoring, navigation routing, personalized trip management. Using only (GSM) fingerprints instead power-hungry GPS WiFi radios, marginal energy consumed zero. This approach non-trivial because we need to process streams highly inaccurate GSM localization (average error over 175 meters) produce accurate trajectory. meets this challenge novel two-pass Hidden Markov Model that sequences directly without converting them geographic coordinates, fuses data low-energy sensors available on commodity smart-phones, accelerometers (to detect movement) magnetic compasses turns). We have implemented Android platform, evaluated it 126 hours (1,074 miles) real driving traces urban environment. find can retrieve 75% user's drive accurately median. An by-product even devices with no or (constituting significant fraction today's phones) contribute benefit data.