relation: https://eprints.uet.vnu.edu.vn/eprints/id/eprint/2796/
title: Enhancing WiFi based Indoor Positioning by Modeling Measurement Data with GMM
creator: Hoang, Manh Kha
creator: Duong, Thi Hang
creator: Vu, Trung Kien
creator: Trinh, Anh Vu
subject: Communications
description: Abstract—Recently, indoor positioning has been investigated by many research groups. Several proposals for indoor positioning have been developed over the last decades, amongst them, WiFi fingerprinting based is considered as the most promissing approach. However, many challenges are still remained for improving the accuracy of indoor mobile object localization. This paper presents a novel approach for enhancement of WiFi fingerprinting based indoor positioning by using Gaussian Mixture Model (GMM) to model the measured Received Signal Strength Index (RSSI) distributions. Since the measured data  varies due to obstacles in indoor environments as well as user directions, the proposed approach is able to model the data distribution more precise than the previous popular approach which utilizes single Gaussian. Simulation results demonstrate the effectiveness of the proposed approach compared to the others.
date: 2017-10-18
type: Conference or Workshop Item
type: NonPeerReviewed
identifier:   Hoang, Manh Kha and Duong, Thi Hang and Vu, Trung Kien and Trinh, Anh Vu  (2017) Enhancing WiFi based Indoor Positioning by Modeling Measurement Data with GMM.  In: International Conference on Advanced Technologies for Communications, 18-20, October, 2017, Quy Nhon, Viet Nam.    (In Press)  
relation: http://atc-conf.org/