Title: Fuzzy-based Wi-Fi localisation with high accuracy using fingerprinting

Authors: Michael Ibrahim; Tarek Nabil; Hassan H. Halawa; Hany M. ElSayed; Ramez M. Daoud; Hassanein H. Amer; Mohamed Ismail; Mostafa Wahish; Mohamed Ibrahim

Addresses: Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt ' Electronics and Communications Engineering Department, The American University in Cairo, P.O. Box 74, AUC Avenue, New Cairo 11835, Egypt

Abstract: In this paper, a localisation system based on Wi-Fi fingerprinting and fuzzy data analysis is presented. Three localisation techniques were used, Euclidean distance, K-nearest neighbours (KNN), and weighted K-nearest neighbours (WKNN), to get three independent estimations of a user's location. Then fuzzy analysis is used to combine the three estimates to achieve highly-accurate localisation. Two experiments were conducted in order to test the proposed new technique and compare it to the traditional fingerprinting techniques present in the literature. The results of the experiments proved that the proposed technique outperforms the traditional techniques.

Keywords: localisation; fingerprinting; fuzzy; Euclidean distance; K-nearest neighbours; KNN; weighted K-nearest neighbours; WKNN.

DOI: 10.1504/IJSCC.2018.088330

International Journal of Systems, Control and Communications, 2018 Vol.9 No.1, pp.1 - 19

Received: 11 Oct 2016
Accepted: 08 Feb 2017
Published online: 04 Dec 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article