Multi-Modal Sensing Residual-Corrected GNN for mmWave Path Loss Prediction via Synesthesia of Machines

TL;DR

A multi-modal sensing residual-corrected graph neural network (MM-ResGNN) framework is proposed for millimeter-wave path loss prediction in vehicular communications for the first time, enabling accurate path loss prediction in unseen vehicular environments with limited labeled data.

Abstract

To support sixth-generation (6G)-enabled intelligent transportation systems (ITSs), a multi-modal sensing residual-corrected graph neural network (MM-ResGNN) framework is proposed for millimeter-wave (mmWave) path loss prediction in vehicular communications for the first time. The propagation environment is formulated as an environment sensing path loss graph (ESPL-Graph), where nodes represent the transmitter (Tx) and receiver (Rx) entities and edges jointly describe Tx--Rx transmission links and Rx--Rx spatial correlation links. Meanwhile, a geometry-driven physical baseline is introduced to decouple deterministic attenuation trends from stochastic residual variations. A vehicular multi-modal path loss dataset (VMMPL) is constructed, which covers three representative scenarios, including the urban wide lane, urban crossroad, and suburban forking road environments, and achieves precise alignment between RGB images and global semantic information in the physical space, and link-level ray-tracing (RT)-based path loss data in the electromagnetic space. In MM-ResGNN, topology-aware graph representations and fine-grained visual semantics are synergistically integrated through a gated fusion mechanism to estimate the path loss residual relative to the physical baseline. Experimental results demonstrate that MM-ResGNN achieves significant improvements over empirical models and conventional data-driven baselines, with a normalized mean squared error (NMSE) of 0.0098, a mean absolute error (MAE) of 5.7991~dB, and a mean absolute percentage error (MAPE) of 5.0498\%. Furthermore, MM-ResGNN exhibits robust cross-scenario generalization through a few-shot fine-tuning strategy, enabling accurate path loss prediction in unseen vehicular environments with limited labeled data.

Figures (10)

• Figure 1: Illustration of the mmWave path loss prediction task in the urban scenario. The Tx (red vehicle) communicates with a dense grid of ground Rxs (blue dots). The prediction targets are the links connected to the nearest 50 Rxs (highlighted in the green zone).
• Figure 2: Visualization of the constructed VMMPL dataset across three scenarios. The VMMPL dataset includes global masks (building and tree), ego-centric RGB images captured by the vehicle, aligned 3D physical environments in AirSim and Wireless InSite, and the corresponding ground-truth path loss heatmaps.
• Figure 3: Vehicle trajectory density heatmaps across the three representative scenarios. The heatmaps illustrate the spatial coverage and diverse mobility patterns, reflecting different traffic densities and roadway topologies.
• Figure 4: Workflow of the data processing and graph construction pipeline. The raw multi-modal data (left) are spatially aligned and transformed into pixel coordinates. A physics-aware feature extraction module (right) utilizes the Bresenham line algorithm to compute blockage features from semantic masks. Finally, a physical baseline model is applied to decouple the deterministic path loss component, generating the graph-structured inputs and residual targets for the neural network.
• Figure 5: Statistical distribution of the path loss across three scenarios. The left panels present the frequency histograms with mean and median markers, while the right panels depict the empirical CDFs, highlighting the distinct electromagnetic scales.