Indoor localization and navigation via Wi-Fi & bluetooth fingerprinting

• Main Author: Eng, Bryan Ze En
• Other Authors: Oh Hong Lye
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: https://hdl.handle.net/10356/156699
• Institution: Nanyang Technological University

Navigational systems have been an integral part of our everyday lives, and with the advancement in technology, Indoor localization (IL) has become a hot topic for research in recent years. There are numerous methodologies for IL, and one of the most popular methodologies is Wi-Fi fingerprinting. In this report, the author would further expand on the methodology by utilizing deep neural networks (DNN) and transfer learning (TL) on top of fingerprinting to build a model that is able to be integrated in an IL application. Apart from Wi-Fi, an experiment was also conducted with Bluetooth Low-level Energy (BLE) beacons for fingerprinting. In addition to conducting experiments on already available public datasets, this project also covers real-life data with data collected in two locations: Singtel Cognitive and Artificial Intelligence Lab for Enterprises (SCALE@NTU), and a museum building complex. After the data collection and pre-processing of data, DNN experiments were conducted on 3 datasets (SCALE@NTU, museum building complex, UJI Indoor Dataset) to evaluate the performance of the DNN models with regards to the data collected. Transfer Learning was also implemented for the UJI Indoor Dataset to compare the accuracy and run-time performance against traditional DNN.