Efficient Rasterization for Outdoor Radio Wave Propagation

Area of Science:

• Computer Science
• Electrical Engineering
• Computational Electromagnetics

Background:

• Global illumination problems are often solved using conventional beam tracing, an efficient GPU-implemented algorithm.
• Radio wave propagation simulation shares conceptual similarities with light transport, suggesting analogous computational approaches.

Purpose of the Study:

• To adapt and apply GPU-accelerated beam tracing for simulating radio wave propagation.
• To develop a custom parallel rasterization pipeline for radio wave simulation.
• To accurately model complex radio channel characteristics for communication system planning.

Main Methods:

• Implementation of a subset of a 3D rasterization pipeline entirely on the GPU.
• Utilizing a custom, parallel rasterization pipeline for beam creation and evaluation.
• Supporting 2D and 3D frame buffers for simulation output.

Main Results:

• The algorithm provides detailed descriptions of radio channel characteristics, including propagation losses and delay spread.
• Simulation results were validated against real-world network measurements.
• The model accounts for diverse propagation environments and adapts to unknown factors like traffic and vegetation.

Conclusions:

• GPU-accelerated beam tracing is a viable and efficient method for radio wave propagation simulation.
• The developed technique offers accurate modeling of complex radio channels essential for mobile network planning.
• The approach demonstrates flexibility in adapting to real-world environmental factors and measurement data.