Open Source OwlPS 1.3: Towards a Reactive Wi-Fi Positioning System Sensitive to Dynamic Changes

Since 2004, our team has been developing an academic positioning system (PS for short) for hostile environments. The techniques involved are mainly Wi-Fi based. While at the very beginning the system architecture was terminal centric, since 2008 it has become infrastructure centric, reducing the application part to being embedded in any Mobile Terminal (MT), focusing on intrusive localization and the illicit use of wireless network. Last year, we presented the OwlPS 1.2 version and the auto-calibration functionality which reduces the RSSI fingerprinting phase to the minimum.This year, the system proposed was enriched with, first of all dynamic changes exploitation, which means not only the last up-to-date RSSI- fingerprinting calibration, but also the instantaneous RSSI variation due mainly to human presence. Therein, the broken line of sight between Wi-Fi Access Points and MTs or the abnormal attenuation of Signal Strength between various references points are taken into account to configure the K-angle-weighted neighborhood proposed algorithm dynamically. Second, the RSSI cartography models the orientation of MT and the relative positioning of human presence. Third a tuning of OwlPS system development kit is performed in off-line phase. It makes use of a 3D dimensioning tool placing N-APs according to GDOPRSSI n-losses criteria and a Wi-Fi-adapted COST-231 propagation model. It also makes use of smartphone Android-based functionalities to calibrate, on demand, AOI reference points.