Channel estimation for Massive MIMO systems aided by intelligent reflecting surface using semi-super resolution GAN

Authors: Mehrdad Momen-Tayefeh, Mehrshad Momen-Tayefeh, Amir Ali Ghafourian Ghahramani, Ali Mohammad Afshin Hemmatyar
Published in: Signal Processing
DOI: [https://doi.org/10.1016/j.sigpro.2024.109710]

Description

This repository contains the implementation and supplementary materials for the paper "Channel Estimation for Massive MIMO Systems Aided by Intelligent Reflecting Surface Using Semi-Super Resolution GAN". The paper introduces a novel Semi-Super Resolution GAN (SSRGAN) for channel estimation in Massive MIMO systems equipped with an Intelligent Reflecting Surface (IRS). The SSRGAN leverages the correlation among adjacent IRS elements to reduce the complexity of channel estimation by turning off some IRS elements during the estimation phase. It effectively predicts the channels for deactivated elements, transforming low-resolution channel data into high-resolution estimations.

The model demonstrated superior performance in comparison to classical and deep learning-based methods, achieving lower NMSE and faster estimation times, even in high-noise environments.

Contents

• Code: Source codes and datasets for implementing SSRGAN.
• Paper: Channel Estimation for Massive MIMO Systems Aided by Intelligent Reflecting Surface Using Semi-Super Resolution GAN.

Citation

If you find this work useful, please cite the paper:

@article{momen2024channel,
  title={Channel estimation for Massive MIMO systems aided by intelligent reflecting surface using semi-super resolution GAN},
  author={Momen-Tayefeh, Mehrdad and Momen-Tayefeh, Mehrshad and Ghahramani, S AmirAli GH and Hemmatyar, Ali Mohammad Afshin},
  journal={Signal Processing},
  pages={109710},
  year={2024},
  publisher={Elsevier}
}

---

This version reflects the content of your paper and is formatted to fit well on GitHub. Let me know if you need further adjustments!