Database calibration algorithms of an indoor positioning system based on the fingerprint method

There is currently, without a doubt, a great need for real time indoor positioning systems. Therefore, a low cost and low power consumption Real Time Indoor Positioning System (RTIPS) that is integrated with a self-developed indoor Geographic Information System (GIS) has been developed by Hsu, et al of National Cheng Kung University (NCKU) in Taiwan. However, RTIPS test bed was initially set up in classroom site that was small and geometrically simple. Once the application location is larger, the challenge we meet is the positioning database calibration because the costs related to time and labor for calibrating a wide area are much higher than those for a smaller site. The purpose of this paper is to achieve a more scalable system for an indoor positioning system; thus, this paper extends study on the database calibration algorithms of a fingerprint positioning algorithm. In the database calibration stage, the collected signal quality might be affected if the positioning space is geometrically complex. This is because of the fact that the signal transmission paths will be more complicated. As a result, we place an emphasis on utilizing two spatial interpolation methods including: 1) the Inverse Distance Weighting (IDW) method; 2) the Kriging method, used to yield a denser database from the raw data measurements. Additionally, another purpose of this paper is to establish a procedure designed to determine the parameters of the theoretical models since most research has mentioned parameters regarding the 'sill' and 'range' estimation of a semi-variogram that usually depend on a trial and error approach, which consumes a great deal of time. Finally, this paper utilizes the Kriging method as an available tool into the RTIPS and also to achieve a more flexible system in application.