Updatable indoor localization based on BLE signal fingerprint

The need for a robust indoor localization solution is mounting in the recent few year, as the IoT technology is growing fast. In this paper, we discuss existing indoor localization approaches, focusing on the signal strength fingerprint approaches, to which belongs the presented study and based on Bluetooth Low Energy signal. The proposed approach employs the Radial Basis Functions (RBF) to create model that describe the real signal strength-position relationship. The present method is fully implemented on a smartphone, namely, offline data collection, model computation and online position estimation. This allow updating the model by collecting more signal data, therefore improving the localization accuracy with small efforts. The presented results shows that updating the model effectively improves the accuracy of estimation.