Abstract

Indoor localization systems are crucial for a range of applications, including navigation, augmented reality, and emergency services. However, achieving centimeter-level accuracy indoors remains a persistent challenge. Existing solutions based on WiFi, Bluetooth Low Energy, or dedicated Ultra-Wideband infrastructure often suffer from high calibration overheads, limited scalability, and significant deployment costs. Despite the growing availability of personal trackers with Ultra-Wideband capabilities, such as Apple AirTags, their use for human localization has remained a largely unexplored research topic.

This thesis addresses this gap by investigating how personal trackers can be repurposed for accurate, calibration-free, and infrastructure-free indoor localization. First, we propose two complementary systems: UbiLoc, which enhances traditional multilateration through AirTag selection and weighting techniques for static setups. Then, we introduce AirLoc, a dynamic GraphSLAM-based system capable of localizing users even when AirTags are mobile.

We evaluate both systems through extensive experiments in two real-world indoor environments. UbiLoc achieves median localization errors between 26 and 31.5 cm, while AirLoc delivers median accuracy between 5 and 5.4 cm, even under dynamic conditions. We also assess system robustness under varying AirTag densities and examine key factors such as battery usage and privacy implications.

Although personal trackers were originally designed for object tracking, these findings demonstrate that personal trackers can form the backbone of scalable and user-centric indoor localization. This opens new opportunities for low-cost, widely deployable localization systems that require no calibration or dedicated infrastructure, paving the way for smarter and more adaptive environments.

School

School of Sciences and Engineering

Department

Computer Science & Engineering Department

Degree Name

MS in Computer Science

Graduation Date

Fall 12-31-2025

Submission Date

9-18-2025

First Advisor

Moustafa Youssef

Second Advisor

Hamada Rizk

Committee Member 1

Sherif Aly

Committee Member 2

Fadel Digham

Extent

81 p.

Document Type

Master's Thesis

Institutional Review Board (IRB) Approval

Approval has been obtained for this item

Disclosure of AI Use

Thesis text drafting; Thesis editing and/or reviewing

Available for download on Saturday, September 18, 2027

Share

COinS