Wave Propagation in Augmented Reality
(2026) EITM01 20261Department of Electrical and Information Technology
- Abstract
- Wireless communication systems like Wi-Fi and 5G rely on radio waves whose behavior is difficult to interpret in indoor environments without spatial context. Raw channel measurements processed by parameter estimation algorithms like Space-Alternating Generalized Expectation-Maximization (SAGE) extract multipath components described by their angle of arrival, propagation delay, and attenuation. However, numerical tables fail to convey the intuitive spatial context of the measured data within the physical environment. While existing tools mostly rely on simulated data, to our knowledge, no system currently combines empirical multipath component data with physical room recordings for immersive Augmented Reality (AR) visualization.
This... (More) - Wireless communication systems like Wi-Fi and 5G rely on radio waves whose behavior is difficult to interpret in indoor environments without spatial context. Raw channel measurements processed by parameter estimation algorithms like Space-Alternating Generalized Expectation-Maximization (SAGE) extract multipath components described by their angle of arrival, propagation delay, and attenuation. However, numerical tables fail to convey the intuitive spatial context of the measured data within the physical environment. While existing tools mostly rely on simulated data, to our knowledge, no system currently combines empirical multipath component data with physical room recordings for immersive Augmented Reality (AR) visualization.
This thesis builds an end-to-end pipeline to address this gap. A Light Detection and Ranging (LiDAR) scan of the room is reconstructed into a 3D mesh. Multipath components are cast against this mesh and geometrically classified as line-of-sight, single-bounce, multi-bounce, or out-of-bounds. These classified frames stream over User Datagram Protocol (UDP) to a Meta Quest 3 headset, where an Unreal Engine 5 application renders the rays in their correct world positions. A dedicated Hypertext Transfer Protocol (HTTP) service handles the initial setup by aligning the LiDAR scan with the headset's scene mesh.
The system successfully transforms empirical measurements into immersive 3D visualizations. Operating within the headset's performance budget, it achieves a reasonable spatial placement accuracy. Ultimately, this provides researchers with a visually plausible, intuitive tool to analyze wireless channels directly within the measured environment. (Less) - Popular Abstract
- Seeing the invisible: Visualizing radio waves in Augmented Reality
Wireless engineers usually study radio waves as tables of numbers. We built an Augmented Reality tool that lets them see the waves instead, drawn as 3D paths through the actual room, viewed through a headset.
Wi-Fi and 5G keep us connected, but the radio waves that carry our data are invisible. When engineers measure how these signals travel through an indoor environment, the result is a table of numbers. Knowing that a signal bounced at 30 degrees after traveling 12 meters tells you little about what happened in the room itself.
Our goal was to bridge that gap by turning the static numbers into a 3D scene a researcher can step inside and explore. The pipeline... (More) - Seeing the invisible: Visualizing radio waves in Augmented Reality
Wireless engineers usually study radio waves as tables of numbers. We built an Augmented Reality tool that lets them see the waves instead, drawn as 3D paths through the actual room, viewed through a headset.
Wi-Fi and 5G keep us connected, but the radio waves that carry our data are invisible. When engineers measure how these signals travel through an indoor environment, the result is a table of numbers. Knowing that a signal bounced at 30 degrees after traveling 12 meters tells you little about what happened in the room itself.
Our goal was to bridge that gap by turning the static numbers into a 3D scene a researcher can step inside and explore. The pipeline takes a 3D laser Light Detection and Ranging (LiDAR) scan of a real room and combines it with the recorded signal data. A computer reconstructs the path each radio wave took, labeling it as line-of-sight, single-bounce, multi-bounce, or out-of-bounds (when no intersection was found). The reconstructed paths are then streamed wirelessly to a Meta Quest 3 headset, which overlays them onto the user's view of the room.
The hardest part was spatial alignment: making a ray that bounced off a real wall look like it is hitting that same wall in the headset. Our first version asked the user to calibrate the room by placing their controllers in the corners, but small placement errors caused the virtual rays to misalign. We replaced this manual step with an automatic one. The headset builds its own 3D mesh of the room in real time, and our software aligns the LiDAR scan with the headset's scene mesh to snap the two coordinate systems together and calculate a transformation matrix. Placement errors dropped to one to five centimeters.
The result is a tool that replaces numerical tables with spatial intuition. Researchers can walk through a measured environment and see exactly where each signal went and what it bounced off, giving them a clearer foundation for designing the next generation of wireless signals. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/student-papers/record/9243953
- author
- Paesschesoone, Wannes and Thys, Tom
- supervisor
- organization
- course
- EITM01 20261
- year
- 2026
- type
- H1 - Master's Degree (One Year)
- subject
- keywords
- Augmented Reality, Wave Propagation, Multipath Components, Distributed MIMO, SAGE Algorithm, Spatial Registration, Mixed Reality
- report number
- LU/LTH-EIT 2026-1181
- language
- English
- id
- 9243953
- date added to LUP
- 2026-06-29 11:30:43
- date last changed
- 2026-06-29 11:30:43
@misc{9243953,
abstract = {{Wireless communication systems like Wi-Fi and 5G rely on radio waves whose behavior is difficult to interpret in indoor environments without spatial context. Raw channel measurements processed by parameter estimation algorithms like Space-Alternating Generalized Expectation-Maximization (SAGE) extract multipath components described by their angle of arrival, propagation delay, and attenuation. However, numerical tables fail to convey the intuitive spatial context of the measured data within the physical environment. While existing tools mostly rely on simulated data, to our knowledge, no system currently combines empirical multipath component data with physical room recordings for immersive Augmented Reality (AR) visualization.
This thesis builds an end-to-end pipeline to address this gap. A Light Detection and Ranging (LiDAR) scan of the room is reconstructed into a 3D mesh. Multipath components are cast against this mesh and geometrically classified as line-of-sight, single-bounce, multi-bounce, or out-of-bounds. These classified frames stream over User Datagram Protocol (UDP) to a Meta Quest 3 headset, where an Unreal Engine 5 application renders the rays in their correct world positions. A dedicated Hypertext Transfer Protocol (HTTP) service handles the initial setup by aligning the LiDAR scan with the headset's scene mesh.
The system successfully transforms empirical measurements into immersive 3D visualizations. Operating within the headset's performance budget, it achieves a reasonable spatial placement accuracy. Ultimately, this provides researchers with a visually plausible, intuitive tool to analyze wireless channels directly within the measured environment.}},
author = {{Paesschesoone, Wannes and Thys, Tom}},
language = {{eng}},
note = {{Student Paper}},
title = {{Wave Propagation in Augmented Reality}},
year = {{2026}},
}