mmDiff: A Noise-Robust Differentiable Ray-Tracing Framework for mmWave Scene Calibration and Channel Prediction
For wireless network planning, this work addresses the practical bottleneck of using imperfect 3D models for mmWave digital twins, enabling more reliable channel prediction from sparse measurements.
The paper proposes mmDiff, a differentiable ray-tracing framework that uses a directional scattering model to make mmWave channel simulation robust to geometric artifacts in 3D reconstructed models. It achieves superior channel prediction accuracy over prior methods on real-world and synthetic datasets.
3D reconstruction techniques such as LiDAR scanning and photogrammetry have made it practical to build detailed geometric models of real-world environments. Such reconstructed models can potentially serve as the foundation for wireless digital twins and support network planning and optimization. The core challenge is that reconstructed models inevitably contain geometric artifacts such as holes and noisy surfaces, and wireless simulation is highly sensitive to such noise. To solve this problem, we propose a differentiable directional scattering model to approximate the noise-sensitive specular reflection. This approximation smoothly distributes reflected power among nearby ray directions, making the simulator inherently robust to local geometric artifacts in the reconstructed model. We prove mathematically that this approximation preserves asymptotic path-gain accuracy. Building on this idea, we propose mmDiff, an end-to-end differentiable framework for calibrating material properties from sparse mmWave measurements and predicting mmWave channels. We evaluate mmDiff on both real-world and synthetic datasets, and demonstrate its superior performance over prior methods using pure specular reflection in noisy reconstructed geometry.