Improvements for 802.11-Based Location Fingerprinting Systems

Position estimation with 802.11 and location fingerprinting has been a topic in research for quite some time already. But there are still some unaddressed issues that are the reason why such systems are not widely used. First, the positioning accuracy still leaves space for improvements. Second, users generally have no information about the quality of the estimated position. Especially in cases where the positioning error is large, the user's trust in the system suffers if he is not notified. Third, most systems that rely on a location fingerprinting approach need a time and effort consuming setup phase in which the training data has to be collected and processed. In this paper, we give an overview of existing possibilities to improve location fingerprinting systems under each of these three aspects. Several alternative solutions for each problem are presented. Finally, a discussion gives an overall picture of the usability of the solutions when considering the system as a whole.