Indoor Localization using Signals of Opportunity

Abstract

This thesis proposes an algorithm to solve the issue of indoor localization using a source, reference receiver and mobile receiver placed in a convex room. The goal is determination of the position of the mobile receiver. The proposed algorithm consists of two localization techniques known from scientific literature: Time Difference of Arrival(TDOA) and Fingerprinting. In this work the choice was made to focus on two following kinds of signals: white noise and human speech. The proposed algorithm has been tested by means of various variables and these experiments have led to the following conclusions: the TDOA part of the algorithm strongly depends on the synchronization between the receivers while the Fingerprinting performs in an optimal manner when a white noise signal is being used as a source. The Fingerprinting part of the proposed algorithm contains four different ways to compare between the audio signal of the database and the audio signal of a real world situation. It appears that the cross - correlation in time domain yields the best results.