Using Machine Learning Techniques and Wi-Fi Signal Strength for Determining Indoor User Location

Abstract

Indoor Positioning System (IPS) can determine someone’s position inside a building. The common method used is implemented by Wi-Fi signal strength analyzing because WLAN/IEEE 802.11 is almost available anywhere and can be easily integrated with a smartphone. However, Wi-Fi access for indoor localization has problems in signal transmission. It is difficult to determine the presence of user indoor location due to the constantly changing Wi-Fi access point signal. In this study, measured signal strength (Receive Signal Strength/RSS) data from several different access points (Aps) in level 1 and 6 of Nusa Putra University. RSS recorded by Wi-Fi netgear and data processing is done using Google Colab. The training data and testing data are processed using the machine learning techniques such as k-Nearest Neighbor (k-NN), Decision Tree and SVM models. The implementation of results with the WLAN method are expected to improve the accuracy values for indoor user locations. k-NN with k=3 has the optimum accuracy (93%) and the smallest error rate (0.15) while SVM has the smallest accuracy (60%) and the largest error rate (0.8).