Home // EMERGING 2025, The Seventeenth International Conference on Emerging Networks and Systems Intelligence // View article
Authors:
Francisco Javier Somolinos-Simón
Adina Murg
Hanli Liu
Carlos Javier Hellín
Josefa Gómez
Abdelhamid Tayebi
Keywords: neural networks; ray tracing; MIMO systems.
Abstract:
Accurate modeling of wireless channels is essential for the design and optimization of next generation communication networks such as 6G. Traditional ray tracing techniques provide physically consistent simulations but suffer from high computational complexity, limiting their scalability and real-time applicability. This work proposes a neural network based surrogate model for ray tracing in complex 3D environments. This approach leverages multilayer perceptrons to predict the interaction of electromagnetic rays with surfaces, estimating critical channel parameters such as gain, time-of-flight, and propagation angles. The model is trained and validated using datasets generated by the Sionna ray tracing engine in both indoor and large urban scenarios. Results demonstrate that the neural surrogate achieves low prediction errors in key metrics and generalizes well across different environments. This neural ray tracing framework offers a scalable, flexible, and efficient alternative to conventional physics based simulators.
Pages: 13 to 19
Copyright: Copyright (c) IARIA, 2025
Publication date: September 28, 2025
Published in: conference
ISSN: 2326-9383
ISBN: 978-1-68558-292-0
Location: Lisbon, Portugal
Dates: from September 28, 2025 to October 2, 2025