Faculty Supervisor

Dr. Tarek El Salti

Date of Defense

Winter 12-1-2020

Program Name

Honours Bachelor of Computer Science (Mobile Computing)

School

Applied Computing

Keywords

Indoor Location Services (ILSs), Fingerprinting, Wi-Fi, Path Loss exponent (PLe), and K-Nearest Neighbor (KNN)

Department

Faculty of Applied Science & Technology (FAST)

Description

A new set of fingerprinting algorithms called Fingerprinting Line-Based Nearest Neighbor (FLBNN) is proposed. Additionally, the new set is compared to other existing Nearest Neighbor-based algorithms in terms of accuracy, precision, and localization time.

Abstract

Localization is of great importance for several fields such as healthcare and security. To achieve localization, GPS technologies are common for outdoor localization but are insufficient for indoor localization. This is because the accuracy and precision of the users’ indoor locations are influenced by many factors (e.g., multipath signal propagations). As a result, the methodologies and technologies for indoor localization services need to remain continuously under development. A related challenge is the time complexity of the methodologies which impacts the performance of the mobile phones’ limited resources. To address these challenges, a new set of fingerprinting algorithms called Fingerprinting Line-Based Nearest Neighbor (FLBNN) is proposed. Furthermore, the new set is compared to other existing Nearest Neighbor-based algorithms. When the deployment of four access points is considered, the FLBNN algorithms outperform several algorithms in terms of accuracy such as Nearest Neighbor version 2, Nearest Neighbor version 4, and Soft-Range-Limited KNN by approximately 17.1%, 7.8%, and 24.1%; respectively. With regards to precision, the new set of algorithms outperforms Path-Loss-Based Fingerprint Localization (PFL) and Dual-Scanned Fingerprint Localization (DFL) by approximately 7.0% and 60.9%; respectively. Moreover, the FLBNN algorithms have a time complexity of O(t * p) where the term t is the number of deployed centroids and the term p is the number of Path Loss exponents. In addition, the new set of algorithms achieves faster run time compared to those for PFL and DFL. As a result, this Thesis improves the cost and reliability of the indoor location services.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.

Share

COinS