Please use this identifier to cite or link to this item: http://hdl.handle.net/10266/4565
Title: Framework of Indoor User Localization Using Wi-Fi Fingerprinting and Visual Clues
Authors: Thind, Ronika
Supervisor: Rani, Rinkle
Keywords: Indoor User Localization,Classification, data pre-processing, data distribution, GPS
Issue Date: 4-Aug-2017
Abstract: Smartphone industry is a steadily growing one, with its development comes various positioning sensors to fulfill the increasing demand for location based services. Most of these services need additional hardware which causes extra cost and lacks in availability to the common masses. Though GPS receiver comes to the rescue of such services with a high accuracy gradient but, has major drawback of signal attenuation in an indoor environment. In this research we are taking advantage of the built in sensors and their mostly available on tap connectivity with the wireless network .The main advantage of the proposed technique is the low cost involved with no extra hardware requirement. But, the cons of the proposed method like multipath fading due to signal attenuation are not avertible and had been worked upon through the course of this research by adding visual database of a user’s position . In the presented thesis we are working with user localization method in an indoor environment using Wi-Fi fingerprinting method and affirming the results obtained with visual clues. Field experiments have been carried out using indoor positioning system in smartphones. The data obtained have been worked upon using machine learning algorithms to increase the accuracy of the results. The positioning of a user has been tested by using the image database of the same user. The accuracy of the results obtained is empirically around 2.4 meters. The advantage of the proposed framework is that by working on the two off the shelf commercial systems and combining their assets we have built a system that provides us with better result.
URI: http://hdl.handle.net/10266/4565
Appears in Collections:Masters Theses@CSED

Files in This Item:
File Description SizeFormat 
801532043(1)- ronika thind.pdf1.99 MBAdobe PDFThumbnail
View/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.