English

Group Equivariant Convolutional Networks for Pathloss Estimation

Networking and Internet Architecture 2025-11-25 v1

Abstract

This paper presents RadioGUNet, a UNet-based deep learning framework for pathloss estimation in wireless communication. Unlike other frameworks, it leverages group equivariant convolutional networks, which are known to increase the expressive capacity of a neural network by allowing the model to generalize to further classes of symmetries, such as rotations and reflections, without the need for data augmentation or data pre-processing. The results of this work are twofold. First, we show that typical UNet-based convolutional models can be easily extended to support group equivariant convolution (g-conv). Secondly, we show that the task of pathloss estimation benefits from such an extension, as the proposed extended model outperforms typical UNet-based models by up to 0.41 dB for a similar number of parameters in the RadioMapSeer dataset. The code is publicly available on the GitHub page: https://github.com/EricssonResearch/radiogunet

Keywords

Cite

@article{arxiv.2511.17841,
  title  = {Group Equivariant Convolutional Networks for Pathloss Estimation},
  author = {Ziyue Yang and Feng Liu and Yifei Jin and Konstantinos Vandikas},
  journal= {arXiv preprint arXiv:2511.17841},
  year   = {2025}
}

Comments

Accepted in IEEE 6GNet 2025

R2 v1 2026-07-01T07:49:51.755Z