UVIndoorLoc-BLE-AoA&IQ Dataset

This repository contains the UVIndoorLoc-BLE-AoA&IQ, a collection of In-phase and Quadrature (I/Q) signal samples and Angle-of-Arrival (AoA) measurements based on the Bluetooth Low Energy (BLE) 5.1 specification. The project focuses on providing a BLE 5.1 AoA dataset by collecting data in three distinct real-world scenarios at the Escola Tècnica Superior d'Enginyeria (ETSE) of the University of Valencia (UV). It was designed to evaluate the impact of anchor topologies and environmental complexity on positioning accuracy. Dataset Features and Format The dataset provides the following parameters for every advertising event, organized into three stages of processing: 1. Raw Data (raw_data/) Individual files per anchor containing the original output from the u-blox firmware: timestamp: Sub-millisecond synchronization reference. ch: BLE channel index used for the transmission. rssi: Received Signal Strength Indicator in dBm. pac: Sequential packet counter. ms: Internal uptime of the NINA-B411 module. az_ublox, el_ublox: Firmware-estimated Azimuth ($\alpha$) and Elevation ($\epsilon$). iq_data: Sequence of 164 values (82 I/Q pairs) captured during the Constant Tone Extension (CTE). 2. Syncronized Data (processed_iq_fusion/) Merged streams from the four anchors using a temporal matching algorithm: timestamp, channel: Shared event identifiers. rssi_AnchorX, az_AnchorX, el_AnchorX, iq_AnchorX: Individual metrics for each of the 4 anchors. ax_AnchorX, ay_AnchorX: Global coordinates of each anchor for the specific topology. config: Topology identifier (Face-to-Face or 45º Inward). 3. Final Dataset (final_dataset/) Cleaned and labeled datasets ready for positioning algorithms and machine learning training: id: Unique identifier for the measurement session. az_ublox_anchorX, el_ublox_anchorX: Cleaned firmware angles. rssi_anchorX, iq_anchorX: Signal metrics for multi-anchor fusion. real_x, real_y, real_z: Precise Ground Truth (GT) coordinates measured via laser distance meter. Experimental Scenarios The dataset distribution is as follows: Scenario Training Points Evaluation Points Total Area 1.a Laboratory 28 18 $59.286\text{ m}^2$ 1.b Corridor 26 21 $135.235\text{ m}^2$ 2. Meeting Room 26 30 $187.392\text{ m}^2$ 3. Auditorium 25 20 $152.170\text{ m}^2$ Hardware Specifications Receivers: 4 u-blox ANT-B10 antenna boards (8-element patch array). Gateways: 4 Raspberry Pi 5 units synchronized via Chrony (NTP). Transmitter: u-blox C211 beacon (100 ms advertising interval).