Indoor Positioning Using BLE Beacons for Care History Collection

Abstract

In order to reduce the positioning errors that occur in BLE beacons positioning and improve the accuracy of indoor positioning to a level that can be used to collect nursing care history, we proposed two methods: one is to use the average of the coordinates calculated by multiple 3-points sets as the final estimated coordinates, and the other is to use only the RSSI from nearby BLE beacons to create fingerprints for position estimation. We implemented the “receiver” software in JavaScript for RSSI observation from the BLE beacons. We conducted an evaluation experiment of the proposed method using the RSSI excluding outliers by k-NN, and confirmed that the accuracy of the proposed method is higher than that of the conventional method and that it achieves the target accuracy. We proposed a method of position estimation using fingerprint with real-time RSSI observed at a fixed point, and confirmed that it exceeded the target accuracy and reached a higher accuracy.