Indoor distance estimated from Bluetooth Low Energy signal strength: Comparison of regression models

Stefano Bertuletti, Andrea Cereatti, Michele Caldara, Michael Galizzi, Ugo Della Croce

2016 · 40 citations · 14 references

Concepts

Abstract

Bluetooth Low Energy (BLE) is a wireless technology for exchanging data, over short distances, designed for the Internet-of-Things era. As widely supported by wearable devices, BLE has the potential to become an alternative for indoor-localization and proximity sensing. The aim of this work was to perform a thorough characterization of the RSSI-distance relationship under controlled conditions using two BLE devices. Four calibration models underwent to a comparative evaluation analysis. The best results were obtained using a polynomial model with a mean distance percentage error equal to 25.7% (0.4 m) in the range 0-3 m. An overall improvement of 14.3% (0.24 m) in the distance estimate compared to the exponential model commonly adopted in the literature was reported.

References

14