althehunter / dl-ce-36

[TWC] Pay Less But Get More: A Dual-Attention-based Channel Estimation Network for Massive MIMO Systems with Low-Density Pilots

This is the official implementation of Pay Less But Get More: A Dual-Attention-based Channel Estimation Network for Massive MIMO Systems with Low-Density Pilots, which has been accepted by IEEE Transactions on Wireless Communications.

To reap the promising benefits of massive multiple-input multiple-output (MIMO) systems, accurate channel state information (CSI) is required through channel estimation. However, due to the complicated wireless propagation environment and large-scale antenna arrays, precise channel estimation for massive MIMO systems is significantly challenging and costs an enormous training overhead. Considerable time-frequency resources are consumed to acquire sufficient accuracy of CSI, which thus severely degrades systems' spectral and energy efficiencies. In this paper, we propose a dual-attention-based channel estimation network (DACEN) to realize accurate channel estimation via low-density pilots, by jointly learning the spatial-temporal domain features of massive MIMO channels with the temporal attention module and the spatial attention module. To further improve the estimation accuracy, we propose a parameter-instance transfer learning approach to transfer the channel knowledge learned from the high-density pilots pre-acquired during the training dataset collection period. Experimental results reveal that the proposed DACEN-based method achieves better channel estimation performance than the existing methods under various pilot-density settings and signal-to-noise ratios. Additionally, with the proposed parameter-instance transfer learning approach, the DACEN-based method achieves additional performance gain, thereby further demonstrating the effectiveness and superiority of the proposed method.

Dataset

Simulation dataset generated with the 3GPP CDL channel model using the Matlab 5G Toolbox. Detailed system setup is referred to Table II of the paper.

Code Usage

• DACEN.py: Module definition of the DACEN
• trainer_from_scratch: Training script to train the DACEN from scratch with low-density pilots
• trainer_TL_source: Training script to train the DACEN from scratch with high-density pilots; the trained DACEN is then used as the source model for parameter transfer
• trainer_TL_target: Training script to train the DACEN with the proposed parameter-instance transfer learning algorithm with low-density pilots (original data samples and generated samples with instance transfer)
• utils: Some utility functions

Citation

If you use this code for your research, please cite our paper:

@article{zhou2023pay,
  title = {Pay Less but Get More: A Dual-Attention-Based Channel Estimation Network for Massive {{MIMO}} Systems with Low-Density Pilots},
  author = {Zhou, Binggui and Yang, Xi and Ma, Shaodan and Gao, Feifei and Yang, Guanghua},
  year = {2023},
  journal = {IEEE Transactions on Wireless Communications},
  pages = {1--1},
  issn = {1558-2248},
  doi = {10.1109/TWC.2023.3329945}
}