Context driven enhancement of RSS-based localization systems

RSS-based indoor localization systems are widely accepted in the literature as one of the less invasive localization technique. In fact, this range-based approach does not require any special hardware and is available in most standard wireless devices. Furthermore, judicious use of RSS has not a significant impact on local power consumption, sensor size, and cost. In front of these interesting characteristics, the performance of the RSS approach is worst with respect to some more invasive ad hoc hardware range-based solutions (such as Angle of Arrival, Time of Arrival etc…). In this paper we propose a localization method that leveraging the context information, such as the knowledge of being in a given room, increases the localization accuracy of RSS-based methods. Performance evaluation is done via real measurements in an office environment composed of three adjacent rooms.