|
|
基于深度学习的IRS辅助MIMO通信系统的CSI压缩及恢复研究
|
Abstract:
智能反射面(IRS, Intelligent Reflecting Surface)因成本低、功耗低、可提升通信质量等优点被广泛研究。在采用正交频分复用作为多载波调制方案的IRS辅助频分双工多输入多输出(MIMO, Multiple input Multiple Output)通信系统中，为了提升系统的系统增益，用户端(UE, User Equipment)需要将多个信道的信道状态信息(CSI, Channel State Information)通过反馈链路发送至基站端(BS, Base Station)。因此，相比于传统的MIMO系统，该系统中CSI的数据量和反馈开销无疑将会是更加巨大的。针对此问题，本文提出了一种基于注意力机制的深度残差网络IARNet (Inception-Attention-Residual-Net)来对大数据量的CSI进行压缩重建。该网络在传统的Inception网络结构上结合了多卷积特征融合、混合注意力机制以及残差等子模块，这种混合结构可以有效地将大数据量的CSI进行压缩重建。仿真结果表明，与现有的2种深度学习网络相比，IARNet在基于热身法的模型训练方案加持下可以显著提高大数据量CSI的重建质量。
Intelligent reflective surfaces (IRS) have been widely studied due to their advantages such as low cost, low power consumption, and ability to improve communication quality. In this paper, an IRS-assisted multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) communication system is considered. In order to improve the performance gain of the system, the user (UE) needs to send its channel state information (CSI) of several channels to the base station (BS) via feedback link. Therefore, the data volume and feedback overhead of CSI in this system will undoubtedly be much huger, as compared to the conventional MIMO systems. To address this problem, we propose an attention-based deep residual network named IARNet (Inception-Attention-Residual-Net) to compress and reconstruct the CSI with large data volume. The IARNet combines several sub-modules based on the traditional Inception network, such as the multi-convolutional feature fusion module, the hybrid attention module, and the residual module, etc. This hybrid structure can effectively compress and reconstruct the CSI of large data volumes. Simulation results show that with the warm-up training scheme, IARNet can significantly improve the reconstruction quality of CSI of large data volumes, as compared to two existing deep learning networks.
| [1] | 朱政宇, 王梓晅, 徐金雷, 王忠勇, 王宁, 郝万明. 智能反射面辅助的未来无线通信: 现状与展望[J/OL]. 航空学报, 1-16. http://kns.cnki.net/kcms/detail/11.1929.v.20210225.1111.008.html, 2021-02-25. |
| [2] | Cui, M., Zhang, G. and Zhang, R. (2019) Secure Wireless Communication via Intelligent Reflecting Surface. IEEE Wireless Communications Letters, 8, 1410-1414. https://doi.org/10.1109/LWC.2019.2919685 |
| [3] | Zhang, D., Wu, Q., Cui, M., Zhang, G. and Niyato, D. (2021) Throughput Maximization for IRS-Assisted Wireless Powered Hybrid NOMA and TDMA. IEEE Wireless Communications Letters, 10, 1944-1948.
https://doi.org/10.1109/LWC.2021.3087495 |
| [4] | Jiang, M., Li, Y., Zhang, G. and Cui, M. (2021) Achievable Rate Region Maximization in Intelligent Reflecting Surfaces-Assisted Interference Channel. IEEE Transactions on Vehicular Technology, 70, 13406-13412.
https://doi.org/10.1109/TVT.2021.3120308 |
| [5] | Li, Y., Jiang, M., Zhang, G. and Cui, M. (2021) Achievable Rate Maximization for Intelligent Reflecting Surface-Assisted Orbital Angular Momentum-Based Communication Systems. IEEE Transactions on Vehicular Technology, 70, 7277-7282. https://doi.org/10.1109/TVT.2021.3089021 |
| [6] | Simonyan, K. and Zisserman, A. (2014) Very Deep Convolutional Networks for Large-Scale Image Recognition. arXiv preprint arXiv:1409.1556. |
| [7] | He, K., Zhang, X., Ren, S. and Sun, J. (2016) Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 770-778.
https://doi.org/10.1109/CVPR.2016.90 |
| [8] | Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. and Wojna, Z. (2015) Rethinking the Inception Architecture for Computer Vision. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, 27-30 June 2016, 2818-2826. https://doi.org/10.1109/CVPR.2016.308 |
| [9] | Szegedy, C., Ioffe, S., Vanhoucke, V. and Alemi, A. (2016) Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning. arXiv preprint arXiv:1602.07261. |
| [10] | Devlin, J., Chang, M.-W., Lee, K. and Toutanova, K. (2018) BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding. arXiv preprint arXiv:1810.04805. |
| [11] | Sun, Y., Wang, S., Li, Y., Feng, S., Tian, H., Wu, H. and Wang, H. (2019) ERNIE 2.0: A Continual Pre-training Framework for Language Understanding. Proceedings of the AAAI Conference on Artificial Intelligence, 34, 8968-8975.
https://doi.org/10.1609/aaai.v34i05.6428 |
| [12] | Li, M., Zuo, W., Gu, S., Zhao, D. and Zhang, D. (2017) Learning Convolutional Networks for Content-weighted Image Compression. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 3214-3223. https://doi.org/10.1109/CVPR.2018.00339 |
| [13] | Kim, D., Kim, M., Kwon, G. and Kim, D.-S. (2019) Progressive Face Super-Resolution via Attention to Facial Landmark. arXiv preprint arXiv:1908.08239. |
| [14] | 于恒, 梅红岩, 许晓明, 贾慧萍. 基于深度学习的图像压缩算法研究综述[J]. 计算机工程与应用, 2020, 56(15): 15-23. |
| [15] | Wen, C., Shih, W. and Jin, S. (2018) Deep Learning for Massive MIMO CSI Feedback. IEEE Wireless Communications Letters, 7, 748-751. https://doi.org/10.1109/LWC.2018.2818160 |
| [16] | Lu, Z., Wang, J. and Song, J. (2020) Multi-Resolution CSI Feedback with Deep Learning in Massive MIMO System. 2020 IEEE International Conference on Communications, Dublin, 7-11 June 2020, 1-6.
https://doi.org/10.1109/ICC40277.2020.9149229 |
| [17] | Yu, X., Li, X., Wu, H. and Bai, Y. (2020) DS-NLCsiNet: Exploiting Non-Local Neural Networks for Massive MIMO CSI Feedback. IEEE Communications Letters, 24, 2790-2794. https://doi.org/10.1109/LCOMM.2020.3019653 |
| [18] | Guo, J., et al. (2020) AI Enabled Wireless Communications with Real Channel Measurements: Channel Feedback. Journal of Communications and Information Networks, 5, 310-317. https://doi.org/10.23919/JCIN.2020.9200894 |
| [19] | 蒋文斌, 彭晶, 叶阁焰. 深度学习自适应学习率算法研究[J]. 华中科技大学学报(自然科学版), 2019, 47(5): 79-83. |
| [20] | Liu, L., et al. (2012) The COST 2100 MIMO channel model: Channel feedback. IEEE Wireless Communications, 19, 92-99. https://doi.org/10.1109/MWC.2012.6393523 |
| [21] | Kingma, D.P. and Ba, J. (2017) Adam: A Method for Stochastic Optimization. arXiv preprint arXiv:1412.6980. |
