Effects of Body Shadowing in LoRa Localization systems

The purpose of this database of measurements with and without body obstruction is to provide the scientific community with a tool to evaluate different localisation methods in these scenarios. The measurements were performed in an outdoor environment with a semi-dense and dense vegetation, located at the Universidad del Norte in Colombia. The data presents the RSSI values transmitted by a LoRa WisTrio RAK5205 node and received by four Risinghf LoRaWAN gateways composing the network, operating at 915 MHz. Additionally, the node and gateways locations in geographic coordinates are provided. Measurements were made by placing the node on a tripod for each test position (scenario without human shadowing - NHS) and then placing the node on a person's chest (scenario with human shadowing - HS) for the same positions.

The database is composed of three groups:
- Raw data:
    RSSI_NHS_raw.csv: Containing 50 packets per position in No Human Shadowing scenario.
    RSSI_HS_raw.csv: Containing 50 packets per position with reception at 3 or 4 gateways, in Human Shadowing scenario.
- Cleaned data: Only for the 16 node positions where the messages were received at the four gateways, after outlier removal (See https://www.mathworks.com/matlabcentral/fileexchange/62954-mean_removing_outliers_tukey-x-rmzerovals?s_tid=mwa_osa_a).
    NHS_16ps.csv: Containing 40 packets per position in No Human Shadowing scenario.
    HS4GWs_16ps.csv: Containing 40 packets per position in Human Shadowing scenario.

Additionally, the geographical coordinates for the gateways are provided in the PosGWs.csv file.

The methodology for data collection and the first results will be published in a research journal.