Sigfox and LoRaWAN Datasets for Fingerprint Localization in Large Urban and Rural Areas

The goal of these LPWAN datasets is to provide the global research community with a benchmark tool to evaluate fingerprint localization algorithms in large outdoor environments with various properties. An identical collection methodology was used for all datasets: during a period of three months, numerous devices which contain a GPS receiver periodically obtained new location data, which was sent to a local data server via a Sigfox or LoRaWAN message. Together with network information such as the receiving time of the message, base station IDs' of all receiving base stations and the Received Signal Strength Indicator (RSSI) per base station, this location data was stored in one of the three LPWAN datasets: 

• lorawan_dataset_antwerp.csv
  • 123 529 LoRaWAN messages, obtained in the city center of Antwerp, Belgium
• sigfox_dataset_antwerp.csv
  • 14 378 Sigfox messages, obtained in the city center of Antwerp, Belgium
• sigfox_dataset_rural.csv
  • 25 638 Sigfox messages, obtained in a rural area between Antwerp, Belgium and Ghent, Belgium

As the rural and urban Sigfox datasets were recorded in adjacent areas, many base stations that are located at the border of these areas can be found in both datasets. However, they do not necessarily share the same identifier: e.g. ‘BS 1’ in the urban Sigfox dataset could be the same base station as ‘BS 36’ in the rural Sigfox dataset. If the user intends to combine both Sigfox datasets, the mapping of the ID's of these base stations can be found in the file:

• sigfox_bs_mapping.csv

The collection methodology of the datasets, and the first results of a basic fingerprinting implementation are documented in the following journal paper:

http://www.mdpi.com/2306-5729/3/2/13