Localization in IEEE 802.11 networks by using the nelder-mead method

Abstract

This papers addresses the problem of localization of mobile devices in WiFi IEEE 802.11 networks considering indoor applications. The key idea of this work is estimate the distance between a mobile receiver and an access point based on the signal propagation losses. Five prediction models for the propagation losses are used: free-space, Recommendation ITU-R P.1238-1, WPS, exponential and potential regression. The Nelder-Mead Method was used to determine the mobile position from its distances to three access points. Experimental results shows that the best localization performance is obtained by using potential regression.