A labeled RF signal dataset for UAV detection and classification in the 2.4GHz band under Wifi and Bluetooth coexistence.

This article presents a labeled radio frequency (RF) dataset for the detection and classification of unmanned aerial vehicles (UAVs) operating in the 2.4 GHz industrial, scientific, and medical (ISM) band. The dataset was acquired using a USRP B210 software-defined radio equipped with bidirectional antennas in an indoor and semi-controlled wireless environment containing realistic WiFi and Bluetooth interference. Data acquisition was performed under multiple operational scenarios, including drone idle, take-off, hovering, movement, and absence of UAV activity. The collected RF signals were transformed into time-frequency spectrogram images using the Short-Time Fourier Transform (STFT) to support machine learning and deep learning applications [1] The dataset was annotated using LabelImg [2] in a YOLO-compatible format to support object detection and classification tasks. This resource is intended for researchers working on RF-based drone detection, wireless spectrum monitoring [3], and deep learning-based signal classification. The dataset enables benchmarking under realistic interference conditions and supports reproducible research in UAV detection systems.