Indoor positioning based on Wi-Fi Fingerprint Technique using Fuzzy K-Nearest Neighbor

Abstract

Indoor positioning system based on Receive Signal Strength Indication (RSSI) from Wireless access equipment have become very popular in recent years. This system is very useful in many applications such as tracking service for older people, mobile robot localization and so on. While Outdoor environment using Global Navigation Satellite System (GNSS) and cellular [14] network works well and widespread for navigator. However, there was a problem with signal propagation from satellites. They cannot be used effectively inside the building areas until a urban environment. In this paper we propose the Wi-Fi Fingerprint Technique using Fuzzy set theory to adaptive Basic K-Nearest Neighbor algorithm to classify the labels of a database system. It was able to improve the accuracy and robustness. The performance of our simple algorithm is evaluated by the experimental results which show that our proposed scheme can achieve a certain level of positioning system accuracy.