Um estudo sobre abordagens para localização indoor

Abstract

This study presents a survey of several existing approaches to indoor localization was performed, creating benchmarks and defining the most appropriate approaches to be implemented at the Federal University of Ceará Quixadá campus. Here, the main criterion was imposed that did not exist to implement change for the environment and the inclusion of any extra equipment. With this technique was chosen based on fingerprint data of the power of Wi-Fi network and the X, Y and Z axes of the magnetic field, because the whole environment is covered by Wi- Fi networks and magnetic field sensors are easily found on some mobile devices. The KNN and K-Means: In addition to implementing the same technique using the power of Wi-Fi antennas and the magnetic field, also for two different data classification algorithms were used. Finally, the testing environment using applications developed for the android platform have been performed. These tests showed poor performance in all algorithms being what was better was that used Wi-Fi as a data source and the K -Means algorithm as classifier had 60 % accuracy.