RobustPathFinder: Handling Uncertainty in Indoor Positioning Techniques

Abstract

Indoor navigation technologies rely on accurately determining the position of a moving person or robot to guide her actively through the building. This cannot be done using GPS because it is indoors. We have been dealing with this problem in the context of designing and building a system for helping blind people navigate in public buildings. However, indoor localization based on RSSI is not very accurate, since the quality of the signal is influenced by many factors that include, for instance, wall reflections or people passing by. The fact that the person is moving also introduces several degrees of uncertainty, including a changing number and placement of beacons in-range. The challenge is therefore to develop mechanisms to handle uncertainty. In this paper we propose and evaluate uncertainty-handling mechanisms