This is a PyTorch implementation of the paper "Physics-Informed Implicit Neural Representation for Wireless Imaging in RIS-Aided ISAC System" in IEEE Transactions on Wireless Communications. Its conference version will be presented in ICC 2026, Glasgow, Scotland, UK, May 2025. Arxiv link: https://arxiv.org/abs/2601.15113

This paper introduces implicit neural representation for wireless imaging and applies it to RIS-aided ISAC systems. Specifically, physics-informed loss functions are formulated based on wireless channel models to impose physical constraints during NN training.

Packages

• python==3.8.0
• pytorch==2.0.0
• numpy==1.24.4
• wandb==0.19.8

Training

The training scripts come with several options. An example for training is:

python train.py --wandb_project 'INRImager'

Testing

Inference is performed during each training epoch. Re-execute the inference process derives the predicted image.

Citation

@article{huang2026physics,
  title={Physics-Informed Implicit Neural Representation for Wireless Imaging in {RIS}-Aided {ISAC} System},  
  author={Huang, Yixuan and Yang, Jie and Wen, Chao-Kai and Jin, Shi},
  journal={IEEE Trans. Wireless Commun.},
  volume={25},
  pages={12341--12357},
  year={Feb. 2025},
  publisher={IEEE}
}

@inproceedings{confhuang2026physics,
  title={Physics-Informed Wireless Imaging with Implicit Neural Representation in {RIS}-Aided {ISAC} System},
  author={Huang, Yixuan and Yang, Jie and Wen, Chao-Kai and Li, Xiao and Jin, Shi},
  booktitle={Proc. Int. Conf. Commun. (ICC)},
  pages={1--6},
  year={May 2026}
}