Photon Splatting: A Physics-Guided Neural Surrogate for Real-Time Wireless Channel Prediction

2025-07-06

AI Summary

To address the challenges of real-time prediction and poor generalizability of wireless channels in complex indoor environments, this paper proposes a geometry-aware, physics-guided neural surrogate model. Methodologically, it models wave propagation using surface-attached virtual light sources (photons) and employs geodesic rasterization to map directional wave features onto the angular domain at the receiver, enabling end-to-end inference of channel impulse responses. The model jointly encodes scene geometry, transmitter location, antenna radiation patterns, and receiver motion—enabling zero-shot generalization to unseen configurations without retraining. Evaluated on a real-world indoor scenario with over one thousand receiver positions, the framework achieves ~30 ms prediction latency, high accuracy, and strong physical interpretability. To our knowledge, this is the first channel prediction framework for 6G network planning and wireless digital twins that simultaneously guarantees real-time performance, cross-configuration generalizability, and strict physical consistency.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Computer Vision: Low Level & Physics-based Vision

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

We present Photon Splatting, a physics-guided neural surrogate model for real-time wireless channel prediction in complex environments. The proposed framework introduces surface-attached virtual sources, referred to as photons, which carry directional wave signatures informed by the scene geometry and transmitter configuration. At runtime, channel impulse responses (CIRs) are predicted by splatting these photons onto the angular domain of the receiver using a geodesic rasterizer. The model is trained to learn a physically grounded representation that maps transmitter-receiver configurations to full channel responses. Once trained, it generalizes to new transmitter positions, antenna beam patterns, and mobile receivers without requiring model retraining. We demonstrate the effectiveness of the framework through a series of experiments, from canonical 3D scenes to a complex indoor cafe with 1,000 receivers. Results show 30 millisecond-level inference latency and accurate CIR predictions across a wide range of configurations. The approach supports real-time adaptability and interpretability, making it a promising candidate for wireless digital twin platforms and future 6G network planning.

Problem

• Real-time wireless channel prediction in complex environments
• Mapping transmitter-receiver configurations to full channel responses
• Generalizing to new transmitter positions without retraining

Innovation

• Physics-guided neural surrogate model
• Surface-attached virtual photon sources
• Geodesic rasterizer for angular splatting