WHIRLS: Wideband High-Resolution Labeling of Signals for Spectrum Sensing Applications

Wideband spectrum sensing and awareness is important to ensure wireless security. Specifically, accurate detection and labeling of active transmissions is necessary for further analysis and classification. In this work, we propose WHIRLS, a digital signal processing method for detecting and labeling RF signals in a wideband spectrum. WHIRLS is able to generate accurate labels with independent time and frequency resolutions, making it flexible in various RF environments. We demonstrate the effectiveness and consistency of WHIRLS in both synthetic and over-the-air dataset and make the code public for future research.