Exploiting Bi-Directional Channel Reciprocity in Deep Learning for Low Rate Massive MIMO CSI Feedback
Exploiting Bi-Directional Channel Reciprocity in Deep Learning for Low Rate Massive MIMO CSI Feedback
复制标题
DOI：
10.1109/lwc.2019.2898662
复制
发表时间：
2019-06-01
影响因子：
6.3
通讯作者：
Ding, Zhi
中科院分区：
文献类型：
--
作者：
Liu, Zhenyu;Zhang, Lin;Ding, Zhi
信道状态信息（CSI）反馈对于多输入多输出（MIMO）无线系统在频分双工模式下实现容量增益非常重要。对于大规模 MIMO 系统，CSI 反馈可能会消耗过多带宽并降低频谱效率。这封信提出了一种基于有限反馈和双向互惠信道特征的基于学习的CSI反馈框架。大规模 MIMO 基站利用可用的上行链路 CSI 来帮助从低速率用户反馈中恢复未知的下行链路 CSI。我们提出了两种深度学习架构 DualNet-MAG 和 DualNet-ABS，以显着减少基于多路径互易性的 CSI 反馈负载。 DualNet-MAG 和 DualNet-ABS 可以分别利用 CSI 系数的实部/虚部的幅度和绝对值的双向相关性。实验结果表明，与基于下行链路的架构相比，我们的架构带来了明显的改进。
Channel state information (CSI) feedback is important for multiple-input multiple-output (MIMO) wireless systems to achieve their capacity gain in frequency division duplex mode. For massive MIMO systems, CSI feedback may consume too much bandwidth and degrade spectrum efficiency. This letter proposes a learning-based CSI feedback framework based on limited feedback and bi-directional reciprocal channel characteristics. The massive MIMO base station exploits the available uplink CSI to help recovering the unknown downlink CSI from low rate user feedback. We propose two deep learning architectures, DualNet-MAG and DualNet-ABS, to significantly reduce the CSI feedback payload based on the multipath reciprocity. DualNet-MAG and DualNet-ABS can exploit the bi-directional correlation of the magnitude and the absolute value of real/imaginary parts of the CSI coefficients, respectively. The experimental results demonstrate that our architectures bring an obvious improvement compared with the downlink-based architecture.