Source-linked AI summary
Enhancing 3GPP Urban Channel Models For Terrestrial-to-Non-Terrestrial Communication
Gerhard Schreiber, Chenrui Sun, Joerg Schaepperle, Le-Hang Nguyen, Thorsten Wild
TL;DR
Existing 3GPP urban NTN channel models have limitations for terrestrial and aerial devices operating above ground, creating a need for enhanced models. The paper combines open-source 3D urban-scene generation with Sionna ray tracing to derive height- and elevation-dependent large-scale parameters, reporting good agreement between analytic functions and ray-tracing data for LOS probability, shadow fading, and clutter loss.
Problem
Existing 3GPP NTN channel models have identified issues and limitations for accurately quantifying terrestrial/aerial-to-NTN communication links in urban environments.
Method
The paper generates 3D urban scenes from OSM data and uses Sionna ray tracing to analyze propagation and derive analytic models for height- and elevation-dependent channel parameters.
Results
The analytic functions show good agreement with ray-tracing data for LOS probability, shadow fading, and clutter loss.
Takeaways & Limitations
The models provide enhanced 3GPP-compliant large-scale channel parameters for terrestrial-to-NTN and aerial-to-NTN urban communication analysis.
Abstract
from arXiv · showhide
We have developed enhanced models for large-scale channel parameters intended for terrestrial-to-non-terrestrial (NTN) communication in Urban environments for supporting devices above ground levels, such as drones. The models are formulated as functions of terminal height above ground and elevation angle, applicable to S-Band and Ka-Band frequencies. To ensure realistic models, an open-source 3D scene generator was utilized to create a diverse set of urban scenes across Europe. Sionna ray-tracing was employed to generate path-gain samples for outdoor terminals, covering a range of heights up to 300 meters and elevation angles up to 90°. Our proposed analytical functions show excellent agreement to ray-tracing data for line-of-sight (LOS) probability, shadow-fading (SF), and clutter-loss (CL). Furthermore, a comparative coupling gain analysis with existing 3GPP models reveals notable differences, particularly caused by smaller CL and higher LOS probabilities from proposed models. Designed for ease of use, these models can be seamlessly integrated into simulation tools, offering a practical solution for researchers and engineers.
I. INTRODUCTION AND PROBLEM STATEMENT
The paper develops enhanced, 3GPP-compliant urban NTN channel models for terrestrial and aerial devices, addressing limitations in existing models. It derives large-scale parameters as functions of terminal height and elevation angle using public 3D-scene and ray-tracing tools.
- UAV applications require reliable, low-latency wireless connectivity, motivating appropriate channel models for combined terrestrial and non-terrestrial networks.
- The paper uses public tools to generate 3D urban scenarios and Sionna ray tracing to derive analytic models for LOS probability, shadow-fading deviation, and clutter-loss.
- The proposed models express large-scale channel parameters as functions of terminal height above ground and elevation angle toward a high-altitude base station.
A. Scene Generation and Ray-Tracing
The processing chain combines open-source 3D urban-scene generation with Sionna ray tracing to produce realistic channel data across diverse environments. Spatial resolution is adapted to capture finer below-rooftop variability and broader above-rooftop trends.
- Twenty urban scenes were generated from locations across Europe using OpenStreetMap building, street, and ground data.Building heights were limited to 22.5 m, and materials were randomly assigned because OSM data lacks material information.
- The scene generator models buildings, streets, and ground, assigning materials as 80% brick-stone, 10% concrete, and 10% marble.
- Sionna ray tracing determines LOS/NLOS conditions and path gains for aerial terminals communicating with a high-altitude base station.
- Approximately 1 m resolution is used for below-rooftop data, while 3–30 m resolutions capture broader above-rooftop variations.
- Different elevation angles are configured by positioning the base station, and ray-tracing outputs undergo statistical analysis to obtain analytic large-scale channel models.
B. Channel Data Analytics
The channel-data analysis classifies terminal positions by propagation condition and estimates LOS probability, shadow fading, and clutter loss as functions of height and elevation. These measurements are then used to construct analytic models.
- LOS probability is computed per location as the ratio of LOS terminals to total terminals for each terminal height and elevation angle, then averaged across locations.
- Actual elevation angles are mapped to the nearest 3GPP-compliant grid angle in Θ = {10, 20, 30, 40, 50, 60, 70, 80, 90}°.The typical mapping difference is within ±2°.
- Total path gain combines the Friis free-space path gain with clutter loss and random shadowing.The Friis contribution depends on link distance and carrier frequency.
- Shadow fading is modeled as zero-mean normal large-scale fluctuation, with SF estimated as the standard deviation of measured path gains by elevation and height.
- Raw shadow-fading and clutter-loss samples are calculated separately for LOS and NLOS conditions before fitting analytic models.
A. Line-of-Sight Probability
The paper models LOS probability as a height- and elevation-dependent function, using a sigmoid below rooftop and a fixed value above rooftop or at 90° elevation.
- A. Line-of-Sight Probability: Below rooftop, for 1.5 m ≤ h ≤ hRT, LOS probability is approximated using a sigmoid function.The model parameters Aθ, Xθ, and Wθ are provided in Table II.
- A. Line-of-Sight Probability: For h > hRT or 90° elevation, the LOS probability is set to 1.
- A. Line-of-Sight Probability: Fig. 2 compares raw S-Band LOS-probability data with the fitted model across elevation angles and terminal heights.
B. Shadow-Fading
Shadow-fading models distinguish LOS and NLOS propagation and below- versus above-rooftop terminal heights. The fitted expressions reproduce the reported raw shadow-fading values while capturing above-rooftop LOS behavior caused by back-reflected rays.
- B. Shadow-Fading: Shadow-fading is modeled separately by propagation condition and by whether the terminal is below or above rooftop.
- B. Shadow-Fading: Above-rooftop LOS shadow-fading uses a Morse-potential-like function of terminal height and rooftop-height difference.The function is defined as M(H1, A1, W1, H2, A2, W2) = A1 · e−W1H1 − A2 · e−W2H2.
- B. Shadow-Fading: Fig. 3 reports a good match between raw and modeled LOS shadow-fading values.
- B. Shadow-Fading: Above-rooftop LOS shadow-fading has a significant positive value caused by back-reflected rays.
- B. Shadow-Fading: For NLOS terminals above rooftop, shadow-fading is equal to 0, and Fig. 4 compares raw with modeled results.
C. Clutter-Loss
Clutter-loss models distinguish LOS and NLOS conditions across below- and above-rooftop heights. The reported fits include small, sign-changing LOS clutter-loss and zero above-rooftop NLOS clutter-loss.
- C. Clutter-Loss: Clutter-loss is modeled separately for LOS and NLOS propagation and for below- versus above-rooftop terminal heights.
- C. Clutter-Loss: LOS clutter-loss is small for every elevation angle and terminal height, so it may be set to 0 as in the current 3GPP model.
- C. Clutter-Loss: Below-rooftop LOS clutter-loss is positive, whereas above-rooftop loss is slightly negative because back-reflected rays contribute to positive path gain.
- C. Clutter-Loss: Below-rooftop NLOS clutter-loss is modeled by combining linear and exponential terms.
- C. Clutter-Loss: Above-rooftop NLOS clutter-loss is equal to 0, as shown in Fig. 6.
IV. COMPARISON TO 3GPP MODELS
The paper evaluates coupling gain as a performance indicator for propagation-model implementation and coverage analysis. Fig. 7 compares 3GPP and proposed models across terminal-height distributions and S- versus Ka-Band.
- IV. COMPARISON TO 3GPP MODELS: Coupling gain CDFs are used to validate propagation-model implementation and conduct baseline radio-coverage analyses.
- IV. COMPARISON TO 3GPP MODELS: Fig. 7 compares coupling gains from 3GPP and proposed models under different terminal-height distributions.
- IV. COMPARISON TO 3GPP MODELS: Solid curves represent S-Band, while dashed curves represent Ka-Band in the coupling-gain comparison.
V. CONCLUSION
The paper develops enhanced 3GPP-compliant urban NTN channel models for aerial terminals, using 3D scene generation and Sionna ray tracing to derive height- and elevation-dependent large-scale parameters. Compared with existing 3GPP NTN models, the proposed models produce significant coupling-gain differences associated primarily with lower clutter loss and higher LOS probabilities.
- The proposed models express LOS probability, shadow fading, and clutter loss as functions of terminal height and elevation angle.
- An open-source tool generated numerous 3D urban scenarios from OSM data, while Sionna ray tracing analyzed LOS and NLOS path gains.
- The resulting analytic expressions target accurate large-scale channel characterization for terrestrial, aerial, and non-terrestrial communication links.
- Coupling-gain evaluations revealed significant differences from existing 3GPP NTN models, primarily because the proposed model has lower clutter-loss values and higher LOS probabilities.
- The model may also apply to in-building terminals, while future work will extend modeling to Dense Urban, Suburban, and Rural scenarios.