Refraction paths not visualized correctly and excessive Line-of-Sight rays

[YutoShibata07]

Description

Hi, I'm currently simulating signal propagation characteristics in a small room (~5m on each side) using Sionna RT. I noticed two major issues related to the visualization of traced paths:

1. Refracted paths are not properly visualized, even though the table material is set to itu_wood or itu_glass, which should support refraction since version 1.0.2.
2. Multiple unexpected "Line-of-Sight" paths are shown, including for paths that seem to involve wall reflections.

The room geometry includes a table placed in the center of the room (see the left panel of the image below). Both the transmitter (tag) and receiver (anchor) are placed on top of this table.

Even when using refractive materials (itu_wood, itu_glass), paths that should pass through the table, reflect off the floor, and reach the receiver are not being visualized.

I also tried changing the orientation of the anchor, but it only affected the amplitude of the CIR and did not result in any change in the traced paths.

Moreover, in the two rightmost plots, many paths that are clearly reflected from the walls are shown in gray and are incorrectly classified as Line-of-Sight.

Image reference

• Left: Room geometry with a table at the center
• Middle: Tracing result with table=itu_wood
• Right: Tracing result with table=itu_glass

Environment

• Sionna version: 1.0.2
• Room size: Approximately 5m × 5m × 5m
• Table materials tested: itu_wood, itu_glass

Expected Behavior

• Paths reflecting off the floor after passing through the table should be visible.
• Reflected paths should not be misclassified as Line-of-Sight.

Any advice or clarification would be greatly appreciated. Thank you very much for your time and support!

[faycalaa]

Hi @YutoShibata07,

• The colors correspond to the effects that spawned the path segments: Grey indicates that the path segments originate from the transmitter, while purple represents the reflected segments, etc.

• The default thickness in Sionna RT is set to 10 cm, which may be too thick for refraction to be significant in your setup. I suggest reducing the thickness of the radio material associated with the object. Note that objects like tables should be geometrically modeled as a single plane (a single sheet of triangles), and the thickness of the radio material assigned to it should be correctly set. When computing the transmitted and reflected fields, the thickness of the traversed object is considered. Please refer to the Radio Material documentation for more details.

• The orientation of the radio devices does not impact the geometry of the paths, only their corresponding channel coefficients through the antenna patterns.

[YutoShibata07]

Hi @faycalaa ,
Thank you for your prompt reply!

I understood the first and last point (about path color and orientation).

Regarding the thickness of the table, I also tested with the material set to wood and the thickness set to 0.001m, but this did not result in any change in behavior. We can see there is no traced path under very thin wood table.

The radio_material properties for each scene_object are printed as follows:

Material: table2, itu_wood
Conductivity: [0.0435797]
Relative permittivity: [1.99]
Thickness: [0.001]

Material: floor, itu_concrete
Conductivity: [0.23469]
Relative permittivity: [5.24]
Thickness: [0.1]

Material: ceiling, itu_ceiling_board
Conductivity: [0.0102675]
Relative permittivity: [1.48]
Thickness: [0.1]

Material: no-name-1, itu_plasterboard
Conductivity: [0.0598713]
Relative permittivity: [2.73]
Thickness: [0.1]

I attached my mitsuba file (demo.zip)

If there are any known solutions or examples where refraction paths have been successfully visualized, I would greatly appreciate it if you could share them.

Thank you again for your time and support!

demo.zip

[faycalaa]

Unfortunately, we cannot download and uncompress zip files.

Could you please provide a minimal reproducer inlined in the GitHub message? It could use one of the built-in scenes from Sionna.