Device-free localization using an ensemble of classifiers with a tapped delay line

This work proposes a novel system for device-free localization over an IEEE 802.11 wireless local area network (WLAN). The proposed system monitors the received signal strength (RSS) transmitted from access points (APs). RSS signals are collected for locating the tracked subject. The subject is located by using an ensemble of classifiers: the support vector machine (SVM) and the Bayesian classifier. Moreover, decisions made by the classifiers at different time units are verified by incorporating the tapped delay line (TDL) architecture in the classifiers. Within a distance error of 1.3 meters and 10 taps, the proposed system achieves a high precision rate of 90.6% when using four access points. Fast, inexpensive, and applicable to any WLAN environment, the proposed system is highly promising for diverse applications.