Deep Learning for Massive MIMO CSI Feedback
Deep Learning for Massive MIMO CSI Feedback
复制标题
DOI：
10.1109/lwc.2018.2818160
复制
发表时间：
2018-10-01
影响因子：
6.3
通讯作者：
Jin, Shi
中科院分区：
文献类型：
--
作者：
Wen, Chao-Kai;Shih, Wan-Ting;Jin, Shi
在分频双工模式下，下行信道状态信息（CSI）需要通过反馈链路发送到基站，才能显示出大规模多输入多输出的潜在增益。然而，这种传输受到过多反馈开销的阻碍。在这封信中，我们使用深度学习技术开发了CsiNet，这是一种新的CSI传感和恢复机制，可以学习有效地利用训练样本中的通道结构。CsiNet学习从CSI到接近最优数量的表示（或码字）的转换，以及从码字到CSI的逆转换。我们通过实验证明，与现有的基于压缩感知（CS）的方法相比，CsiNet可以显著提高CSI的重建质量。即使在过低的压缩区域，基于cs的方法不能工作，CsiNet保持有效的波束形成增益。
In frequency division duplex mode, the downlink channel state information (CSI) should be sent to the base station through feedback links so that the potential gains of a massive multiple-input multiple-output can be exhibited. However, such a transmission is hindered by excessive feedback overhead. In this letter, we use deep learning technology to develop CsiNet, a novel CSI sensing and recovery mechanism that learns to effectively use channel structure from training samples. CsiNet learns a transformation from CSI to a near-optimal number of representations (or codewords) and an inverse transformation from codewords to CSI. We perform experiments to demonstrate that CsiNet can recover CSI with significantly improved reconstruction quality compared with existing compressive sensing (CS)-based methods. Even at excessively low compression regions where CS-based methods cannot work, CsiNet retains effective beamforming gain.