WiFi Fingerprinting Dataset for IP-TAG
Indoor location-based services have high requirements for positioning accuracy. Fingerprint positioning methods are popular, where Received Signal Strength (RSS) of WiFi is widely used because of its availability. Our dataset is from the dataset provided in the literature . The WiFi measurements were collected in an area among the bookshelves in a wing of a university’s library building. The collection process was finished with a Samsung Galaxy S3 smartphone and software explicitly developed, and a total of 448 Access Points (APs) were detected during the experiment. We chose 24 collection positions (14 reference points and 10 test points) and the corresponding 64 AP measurements in our experiments to form the dataset. Therefore, the dataset contains 2-dimensional coordinates of each point and 64-dimensional RSS vectors measured at each point. To improve data stability and facilitate the model training, we Gaussian filtered and normalized the RSS vectors.
 G. M. Mendoza-Silva, P. Richter, J. Torres-Sospedra, et al., “Long-term WiFi fingerprinting dataset for research on robust indoor positioning,” Data, vol. 3, no. 1, Art.no. 3, Jan. 2018.
The dataset Folder name “WiFi Fingerprinting Dataset” includes: Fingerprint.csv, 14_train.csv, 10_test.csv.
In the “Fingerprint.csv”, the first two columns of data indicate the coordinates of each point and the remaining columns indicate the RSS measured at each point.
“14_train.csv” and “10_test.csv” are the training set and the testing set, respectively, they can be obtained by processing “Fingerprint.csv”.