WiFi RSS measurements in Tampere University multi-building campus, 2017

This package contains two Wi-Fi RSSI Datasets, namely TIE1 and SAH1.

• TIE1 was collected in the Tietotalo building from Tampere University during the period of August-December 2017 and it is composed of four files:
  • TIE1_training_rss.csv - Radio map or reference dataset as a matrix with 10633 reference samples (rows) and 613 APs (columns) 
  • TIE1_training_coordinates.csv - Ground truth location of the 10633 reference samples (rows) as X,Y,Z and floor being the coordinates in a local system
  • TIE1_test_rss.csv - Operational fingerprints as a matrix with 50 evaluation samples (rows) and 613 APs (columns) 
  • TIE1_test_coordinates.csv - Ground truth location of the 50 evaluation samples (rows) as X,Y,Z and floor being the coordinates in a local system
• SAH1 was collected in the Sahkotalo building from Tampere University during the period of October-December 2017 and it is composed of four files:
  • SAH1_training_rss.csv - Radio map or reference dataset as a matrix with 9291 reference samples (rows) and 775 APs (columns) 
  • SAH1_training_coordinates.csv - Ground truth location of the 9291 reference samples (rows) as X,Y,Z and floor being the coordinates in a local system
  • SAH1_test_rss.csv - Operational fingerprints as a matrix with 156 evaluation samples (rows) and 775 APs (columns) 
  • SAH1_test_coordinates.csv - Ground truth location of the 156 evaluation samples (rows) as X,Y,Z and floor being the coordinates in a local system

Additional considerations:

• All the RSSI values are stored are they were registered with a regular smartphone. In order to generate the fingerprint vectors, the non-detected APs where filled with the default bogus value +100dBm. i.e., the value +100dBm in the training and test rss files indicates that the AP was not detected. Generally, this default bogus value has been replaced with a low RSSI value in experiments, such as -100dBm, -110dBm, -150dBm or -200dBm.
• The coordinates X,Y,Z are expressed in meters in a local coordinate system. Additionally, the sequential floor identifier is also provided.

Please cite the source at Zenodo when using any of these two datasets:

Lohan, E.S; Torres-Sospedra, J.; Gonzalez, A.; "WiFi RSS measurements in Tampere University multi-building campus, 2017", Zenodo, 2021.
https://zenodo.org/record/5174851
https://www.doi.org/10.5281/zenodo.5174851