Environment-Aware Channel Prediction for Vehicular Communications: A Multimodal Visual Feature Fusion Framework

2026-04-02

AI Summary

This work addresses the challenge of achieving high reliability and ultra-low latency in 6G vehicular networks by proposing an environment-aware channel prediction framework based on multimodal visual feature fusion. To overcome the trade-off among accuracy, generalization, and deployability in conventional channel prediction models, the proposed approach uniquely integrates semantic, depth, and positional information into V2I channel modeling. A three-branch neural network extracts multimodal features from panoramic RGB images, GPS data, and other sources, while a squeeze-and-excitation attention gating mechanism enables adaptive feature fusion. Additionally, a dedicated head architecture and a composite multi-constraint loss function are introduced for regression of 360-dimensional angular power spectra. Evaluated on an urban real-world dataset, the method achieves RMSEs of 3.26 dB for path loss, 37.66 ns for delay spread, and 5.05°/5.08° for azimuth/elevation spreads, with mean and median cosine similarities of 0.9342 and 0.9571 for angular power spectra, respectively, demonstrating significantly enhanced prediction accuracy and generalization capability.

Abstract

The deep integration of communication with intelligence and sensing, as a defining vision of 6G, renders environment-aware channel prediction a key enabling technology. As a representative 6G application, vehicular communications require accurate and forward-looking channel prediction under stringent reliability, latency, and adaptability demands. Traditional empirical and deterministic models remain limited in balancing accuracy, generalization, and deployability, while the growing availability of onboard and roadside sensing devices offers a promising source of environmental priors. This paper proposes an environment-aware channel prediction framework based on multimodal visual feature fusion. Using GPS data and vehicle-side panoramic RGB images, together with semantic segmentation and depth estimation, the framework extracts semantic, depth, and position features through a three-branch architecture and performs adaptive multimodal fusion via a squeeze-excitation attention gating module. For 360-dimensional angular power spectrum (APS) prediction, a dedicated regression head and a composite multi-constraint loss are further designed. As a result, joint prediction of path loss (PL), delay spread (DS), azimuth spread of arrival (ASA), azimuth spread of departure (ASD), and APS is achieved. Experiments on a synchronized urban V2I measurement dataset yield the best root mean square error (RMSE) of 3.26 dB for PL, RMSEs of 37.66 ns, 5.05 degrees, and 5.08 degrees for DS, ASA, and ASD, respectively, and mean/median APS cosine similarities of 0.9342/0.9571, demonstrating strong accuracy, generalization, and practical potential for intelligent channel prediction in 6G vehicular communications.

Problem

• environment-aware channel prediction
• vehicular communications
• 6G
• multimodal fusion
• channel state information

Innovation

• multimodal fusion
• environment-aware channel prediction
• visual feature extraction
• angular power spectrum
• squeeze-excitation attention