Indoor localization by WiFi

Abstract
The demand for accurate indoor localization techniques is increasing. The existing infrastructure of 802.11 WiFi networks can be exploited to position a device in a building. In this paper a positioning system is presented to locate a robot device in the Faculty of Computer Science building. Several techniques are experimented and the algorithms K Nearest-Neighbours (KNN), weighted KNN, weighted Centroid, and Gaussian kernel methods are implemented and compared.