RadTwin: Generalizable Wireless Digital Twin for Dynamic Environments

2026-04-25

AI Summary

This work addresses the limited generalization of existing wireless digital twin methods in dynamic environments, where ray tracing relies on precise 3D models and neural radiance fields are confined to static scenes requiring retraining. To overcome these limitations, we propose RadTwin, a novel framework that uniquely integrates point-cloud-driven scene representation, physics-guided sparse attention, and a neural propagation decoder. By explicitly conditioning on scene geometry, RadTwin adapts to dynamic environments without retraining. Experimental results demonstrate that RadTwin substantially outperforms baseline methods in indoor dynamic scenarios, achieving a 31.6% improvement in SSIM (0.846 vs. 0.643) and a 91.96% reduction in LPIPS (0.023 vs. 0.286), highlighting its superior cross-scale generalization capability and data efficiency.

Technology Category

• Computer Vision: Low Level & Physics-based Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Activity and Plan Recognition

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Precisely modeling radio propagation in dynamic wireless environments is fundamental to the realization of wireless digital twins. Traditional ray tracing methods rely on accurate 3D models with detailed environment parameters, while recent neural radiance field approaches learn representations tied to specific static scenes, requiring retraining when environments change. In this paper, we propose RadTwin, a generalizable wireless digital twin framework that explicitly conditions on scene geometry, enabling adaptation to dynamic environments without retraining. RadTwin comprises three key components: 1) a scenario representation network that extracts high-level latent scene features from point clouds, 2) an electromagnetic ray tracing module that computes physics-informed sparse attention masks identifying voxels that physically contribute signals toward each query direction, and 3) a neural propagation decoder that aggregates relevant scene features through masked cross-attention to learn how radio propagation behaves within the given scene geometry. We evaluate RadTwin on a customized dataset of indoor scenes with varying furniture arrangements. Experimental results show that RadTwin achieves 31.6% higher SSIM (0.846 vs. 0.643) and 91.96% lower LPIPS (0.023 vs. 0.286) compared to NeRF2. RadTwin further demonstrates superior cross-scale performance and high generalization and data efficiency, representing a significant advancement toward practical digital network twins for dynamic wireless environments.

Problem

• wireless digital twin
• dynamic environments
• radio propagation
• generalizable modeling
• scene geometry

Innovation

• wireless digital twin
• generalizable modeling
• scene-conditioned neural propagation
• physics-informed attention
• dynamic environment adaptation