当前位置: X-MOL 学术IEEE Wirel. Commun. Lett. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
DARENet: Data ARrangEment Neural Network for Eigenvector-Based CSI Feedback
IEEE Wireless Communications Letters ( IF 4.6 ) Pub Date : 5-31-2024 , DOI: 10.1109/lwc.2024.3407813
Ruofei Gao 1 , Xiaotao Li 1 , Wai Chen 1
Affiliation  

In 5G communications, the precision of Channel State Information (CSI) feedback is vital, and the massive Multiple-Input Multiple-Output (MIMO) systems rely heavily on this for optimal performance. While eigenvector-based methods using Deep Learning (DL) have innovated CSI feedback mechanisms, they do not fully exploit the intrinsic correlations within CSI that are instrumental for feedback optimization. To bridge this gap, we propose the Data ARrangEment Neural Network (DARENet), a Convolutional Neural Network (CNN) based framework that effectively utilizes the inherent correlations present in CSI eigenvectors with cross-polarized antennas. DARENet’s capabilities are validated through rigorous testing on 5 public datasets, where it consistently outperforms the established EVCsiNet and PolarDenseNet in terms of recovery performance, computational efficiency, and model complexity.

中文翻译:


DARENet:用于基于特征向量的 CSI 反馈的数据排列神经网络



在 5G 通信中,信道状态信息 (CSI) 反馈的精度至关重要,大规模多输入多输出 (MIMO) 系统在很大程度上依赖于此来实现最佳性能。虽然使用深度学习 (DL) 的基于特征向量的方法创新了 CSI 反馈机制,但它们没有充分利用 CSI 中有助于反馈优化的内在相关性。为了弥补这一差距,我们提出了数据排列神经网络(DARENet),这是一种基于卷积神经网络(CNN)的框架,可有效利用交叉极化天线的 CSI 特征向量中存在的固有相关性。 DARENet 的功能通过对 5 个公共数据集的严格测试进行了验证,在恢复性能、计算效率和模型复杂性方面始终优于现有的 EVCsiNet 和 PolarDenseNet。
更新日期:2024-08-19
down
wechat
bug