Joshua Riojas

Class of 2026

Major: Electrical Engineering

Mentor: Ben Abbot, PhD

St. Mary’s University

Comprehensive Signal Strength Mapping for Indoor Object Localization

Congested environments resulting in numerous reflections from one or more radio frequency

(RF) sources exacerbate the accuracy of Time Space Positioning Information (TSPI). The St.

Mary’s Unmanned Aerial Systems (UAS) Lab, being a highly reflective building (almost entirely

metal), renders the use of GPS signals for indoor localization impractical. Consequently, this has

led to exploring the utilization of RF reflections to determine an object’s position. Recently,

Kimberly Tse, a graduate student from St. Mary’s University, designed a Convolutional Neural

Network (CNN)-based TSPI localization model, achieving a 94% accuracy with synthetic data

simulated via MATLAB and validated by real-world signal strengths gathered across a small

area of the UAS Lab [1]. This paper presents a different approach to gathering signal strengths

across the UAS Lab to provide comprehensive data for enhancing the machine learning model’s

localization accuracy. We utilized a calibrated in- frared camera system with real-time TSPI to

gather accurate positioning truth data and employed robotic cars to cover a specified area,

thereby laying the groundwork for future analysis and model training with submillimeter

precision.

Keywords: Radio frequency, Signal strengths, Indoor localization, Highly reflective

environments, Synthetic data