Localização em Ambientes Internos utilizando redes IEEE 802.11 e algoritmo WKNN

Abstract

In recent years, with the evolution of information and communication technologies, new techniques for locating objects or users in indoor environments have emerged. Despite this, indoor location systems (ILSs) continue to be a challenge since some proposals depend on sensors with specific technology and high cost to obtain adequate system accuracy. In addition, many existing ILS proposals have limitations in their Human- Machine Interface (HMI), making operation and usability difficult. In this work we present a high computational capacity, easy operation and simple usability system. The proposed ILS determines the location of an user, from his smartphone, using signatures of preexisting Wi-Fi access points in the environment, a database created in an off-line phase and WKNN (Weighted K-Nearest Neighbor) supervised learning algorithm. The proposed ILS showed a RMSE (Root Mean Square Error) location error of 2,87m, surpassing recent work in the literature, where the lowest RMSE location error obtained was 4,31m. In addition, our proposal provides HMI configuration facilities.