Indoor UWB Jammer Localization Dataset (Classification, Regression, Domain Alignment)

This dataset is used in the paper "Machine and Deep Learning for Indoor UWB Jammer Localization". It contains measurements of two types: per-corner diagnostic readings from DW-3000 registers and raw Channel Impulse Response (CIR) taps. The dataset is organized into two parts: Dataset.zip contains the source domain data in four folders (1, 2, 3, 4), each corresponding to a specific receiver position. This part has a total of 461,795 samples. test_setup.zip contains measurements from a modified room layout for testing, also in four folders (1, 2, 3, 4). This part has a total of 28,793 samples. The Positions.xlsx file, located inside the Dataset folder, provides the (x,y) coordinates for all samples. All data points are labeled with their folder name, identifying the receiver that recorded the data. This setup allows the dataset to be used for both classification (assigning the jammer to a predefined position on the grid) and regression (precise jammer localization). See Additional details for Repository URL.