RadioVIL: Anomaly-Aware Diffusion Models for Radio Map Inpainting and Zero-Shot Vehicle Localization

2026-08-17

AI Summary

This study addresses the challenges of over-smoothing and loss of vehicle scattering features in radio map reconstruction by proposing an anomaly-aware diffusion framework that formulates completion and localization as a physical inverse problem. The proposed DMILO algorithm hierarchically isolates scattering anomalies and incorporates L1 sparse bias optimization to preserve authentic physical textures while enabling zero-shot localization. Experimental results demonstrate that this method achieves an LPIPS score of 0.0587, a zero-shot localization recall rate of 75.20%, and a mean localization error of 3.31 meters. These metrics significantly outperform existing baselines, effectively resolving the dual challenges of high-fidelity map completion and training-free localization in complex vehicular environments.

Abstract

High-precision radio map construction is essential for emerging 6G Integrated Sensing and Communication (ISAC) applications, including digital twins and intelligent transportation. However, existing deep learning methods predominantly treat this as a pure image completion task, resulting in over-smoothed reconstructions that fundamentally erase high-frequency scattering signatures of dynamic physical entities such as hidden vehicles. To overcome this, we propose RadioVIL, an efficient two-stage framework that reformulates joint radio map inpainting and zero-shot vehicle localization as a prior-guided physical inverse problem. Specifically, we first train a Denoising Diffusion Probabilistic Model (DDPM) to capture the structural generative prior of the environment. During inference from highly sparse measurements, we employ a Diffusion-based Mediating Intermediate Layer Optimization (DMILO) algorithm. By optimizing an L1-regularized sparse deviation term, DMILO mathematically isolates vehicle scattering anomalies layer-by-layer without unfolding the entire denoising chain. Extensive experiments demonstrate that while conventional reconstruction baselines fail to detect hidden vehicles, and the zero-shot diffusion baseline achieves only limited detection ability due to forced semantic harmonization, RadioVIL preserves authentic physical textures, yielding the best LPIPS of 0.0587 in our evaluation. Uniquely, it unlocks accurate zero-shot vehicle localization directly from sparse radio maps, securing a 75.20% Recall and a 3.31-meter average error, paving a robust way for ISAC at the 6G edge.

Problem

• Radio Map Inpainting
• Zero-Shot Vehicle Localization
• 6G ISAC
• High-Frequency Scattering Signatures
• Sparse Measurements

Innovation

• Radio Map Inpainting
• Zero-Shot Vehicle Localization
• Diffusion Models
• DMILO Algorithm
• Physical Inverse Problem