A novel radio map construction method to reduce collection effort for indoor localization

Indoor localization using the fingerprinting technique, namely, the radio map, has attracted much attention in the research community recently. However, constructing a complete radio map is extremely labor-extensive and time-consuming, especially for a wide area. Although some works have been done to reduce the number of calibration points, the accuracy decreases if there are not enough fingerprints. In this paper, we propose a novel method based on the radio propagation model to construct a radio map with full fingerprints. In the radio map, the calibration points (CPs), i.e., fingerprints, are classified into two categories: the primary CPs chosen to collect received signal strength (RSS) artificially, and the secondary CPs obtained through some calculations. Based on the radio map constructed, we employ the Voronoi diagram to divide the area into several Voronoi regions and restrict the localization algorithm to run in a specific Voronoi region to reduce computational complexity. The comparison results show that our method saves a lot of time and human effort in collecting RSS samples, and achieves much higher accuracy than other existing schemes.