Indoor Localization using Voronoi Tessellation

Abstract

Recently the use of received signal strength values from a wireless local area network has received significant research interest for indoor localization. This work investigates a Voronoi-based interpolation method to improve indoor localization performance. The region of interest is spanned by reference measurement locations, termed as anchors. The proposed method is shown to outperform well-known localization techniques such as the k-Nearest Neighbor (k-NN) and the Inverse Distance Weighting (IDW) methods in terms of accuracy and precision. Our results show that for a 20 m x 20 m room the proposed scheme can achieve a location accuracy of 5.7 m with at most 5 anchors, whereas the IDW and k-NN techniques attain location accuracies of only 6.1 m and 6.5 m, respectively, under the same conditions. These performance gains are achieved while maintaining the same number of anchors in the system calibration phase for all the considered techniques.

Keywords

• indoor environments
• interpolation
• radio propagation
• simultaneous localization and mapping
• wireless LAN