Date of Award
Daniel Felix Ritchie School of Engineering and Computer Science, Electrical and Computer Engineering
Mohammad H. Mahoor
Computer engineering, Artificial intelligence, Global navigation system
Outdoor positioning systems based on the Global Navigation Satellite System have several shortcomings that have deemed their use for indoor positioning impractical. Location fingerprinting, which utilizes machine learning, has emerged as a viable method and solution for indoor positioning due to its simple concept and accurate performance. In the past, shallow learning algorithms were traditionally used in location fingerprinting. Recently, the research community started utilizing deep learning methods for fingerprinting after witnessing the great success and superiority these methods have over traditional/shallow machine learning algorithms. The contribution of this dissertation is fourfold:
First, a Convolutional Neural Network (CNN)-based method for localizing a smartwatch indoors using geomagnetic field measurements is presented. The proposed method was tested on real world data in an indoor environment composed of three corridors of different lengths and three rooms of different sizes. Experimental results show a promising location classification accuracy of 97.77% with a mean localization error of 0.14 meter (m).
Second, a method that makes use of cellular signals emitting from a serving eNodeB to provide symbolic indoor positioning is presented. The proposed method utilizes Denoising Autoencoders (DAEs) to mitigate the effects of cellular signal loss. The proposed method was evaluated using real-world data collected from two different smartphones inside a representative apartment of eight symbolic spaces. Experimental results verify that the proposed method outperforms conventional symbolic indoor positioning techniques in various performance metrics.
Third, an investigation is conducted to determine whether Variational Autoencoders (VAEs) and Conditional Variational Autoencoders (CVAEs) are able to learn the distribution of the minority symbolic spaces, for a highly imbalanced fingerprinting dataset, so as to generate synthetic fingerprints that promote enhancements in a classifier's performance. Experimental results show that this is indeed the case. By using various performance evaluation metrics, the achieved results are compared to those obtained by two state-of-the-art oversampling methods known as Synthetic Minority Oversampling TEchnique (SMOTE) and ADAptive SYNthetic (ADASYN) sampling.
Fourth, a novel dataset of outdoor location fingerprints is presented. The proposed dataset, named OutFin, addresses the lack of publicly available datasets that researchers can use to develop, evaluate, and compare fingerprint-based positioning solutions which can constitute a high entry barrier for studies. OutFin is comprised of diverse data types such as WiFi, Bluetooth, and cellular signal strengths, in addition to measurements from various sensors including the magnetometer, accelerometer, gyroscope, barometer, and ambient light sensor. The collection area spanned four dispersed sites with a total of 122 Reference Points (RPs). Before OutFin was made available to the public, several experiments were conducted to validate its technical quality.
Copyright is held by the author. User is responsible for all copyright compliance.
Received from ProQuest
Alhomayani, Fahad, "Deep Learning Methods for Fingerprint-Based Indoor and Outdoor Positioning" (2021). Electronic Theses and Dissertations. 1887.
Computer engineering, Artificial intelligence