BluePIL: a Bluetooth-based PassIve Localization Method

Indoor location services are of great relevance for several applications ranging from mobile marketing and navigation to healthcare technologies. Wireless technologies provide a method of gauging performance indicators through the capturing of signals emitted by mobile devices. Since the Bluetooth technology has seen a steady growth over the years, providing a viable alternative to 802.11 protocols, this paper presents BluePIL, a fully passive system for Bluetooth device identification and localization designed as a distributed streaming architecture delivering results in near-real-time. This approach enables a contact and activity tracing with different data privacy dimensions. BluePIL relies on parts of the Bluetooth address for device identification and a modified multi-lateration algorithm using a path loss model for device localization. The results were obtained using Ubertooth Bluetooth sensors and low-cost hardware (ASUS Tinkerboard devices), showing that the approach achieves localization accuracies of 1 m to 1.4 m within a space of around 12 $\mathrm{m}^{2}$ and 25 $\mathrm{m}^{2}$, respectively.