Recovering the invisible signals for urban traffic monitoring using distributed acoustic sensing

Distributed acoustic sensing (DAS) is an emerging technology for recording vibration signals via the optical fibers buried in subsurface conduits. The capability of DAS to record seismic signals in the oil and gas industry or natural earthquakes has been well explored. However, as an important part of the city infrastructure, the usage of optical fibers has drawn relatively less attention aside from its functionality as a telecommunication cable. Here, we focus on the application potential of DAS data in the urban environment, particularly its capability for traffic monitoring. We propose a sophisticated workflow for processing the urban DAS data, including bandpass filtering, removing bad traces, amplitude scaling, and local F-K filtering (thresholding and sector). Considering the signal leakage after these processing steps, we propose to apply a recently developed signal-retrieving approach based on a dictionary learning scheme in a secondary stage. Despite the raw DAS data’s very low signal-to-noise ratio (SNR), the resulting signals are much cleaner and high-quality.