file.pdf (77.35 kB)
Download fileA Learning Algorithm for Localizing People Based onWireless Signal Strength that Uses Labeled and Unlabeled Data
journal contribution
posted on 2006-07-01, 00:00 authored by Mary Berna, Brennan Sellner, Brad Lisien, Sebastian Thrun, Geoffrey J. Gordon, Frank PfenningThis paper summarizes a probabilistic approach for
localizing people through the signal strengths of a
wireless IEEE 802.11b network. Our approach uses
data labeled by ground truth position to learn a probabilistic
mapping from locations to wireless signals,
represented by piecewise linear Gaussians. It then
uses sequences of wireless signal data (without position
labels) to acquire motion models of individual
people, which further improves the localization
accuracy. The approach has been implemented and
evaluated in an office environment.