Differentiable and Learnable Wireless Simulation with Geometric Transformers

TL;DR

Wi-GATr is a novel neural surrogate model for wireless signal propagation that uses geometric transformers to accurately and efficiently predict channel observations from scene primitives, outperforming traditional methods.

Contribution

The paper introduces Wi-GATr, a learnable geometric transformer-based model for wireless simulation that improves accuracy and efficiency over existing methods.

Findings

Wi-GATr achieves over 35% lower error than hybrid techniques.

Wi-GATr is accurate, fast, and sample-efficient.

It generalizes well to real-world scenarios.

Abstract

Modelling the propagation of electromagnetic wireless signals is critical for designing modern communication systems. Wireless ray tracing simulators model signal propagation based on the 3D geometry and other scene parameters, but their accuracy is fundamentally limited by underlying modelling assumptions and correctness of parameters. In this work, we introduce Wi-GATr, a fully-learnable neural simulation surrogate designed to predict the channel observations based on scene primitives (e.g., surface mesh, antenna position and orientation). Recognizing the inherently geometric nature of these primitives, Wi-GATr leverages an equivariant Geometric Algebra Transformer that operates on a tokenizer specifically tailored for wireless simulation. We evaluate our approach on a range of tasks (i.e., signal strength and delay spread prediction, receiver localization, and geometry reconstruction) and find that Wi-GATr is accurate, fast, sample-efficient, and robust to symmetry-induced transformations. Remarkably, we find our results also translate well to the real world: Wi-GATr demonstrates more than 35% lower error than hybrid techniques, and 70% lower error than a calibrated wireless tracer.