Multi-resolution CSI Feedback with Deep Learningin Massive MIMO System-笔记

In this paper, a novel feedback network named
CRNet is proposed to achieve better performance via extracting
CSI features on multiple resolutions. An advanced training
scheme that further boosts the network performance is also
introduced. Simulation results show that the proposed CRNet
outperforms the state-of-the-art CsiNet under the same computa-
tional complexity without any extra information.（本文提出了一种新的反馈网络CRNet，通过提取多种分辨率下的CSI特征来获得更好的性能。还介绍了进一步提高网络性能的高级训练方案。仿真结果表明，在不需要额外信息的情况下，在相同的计算复杂度下，所提出的CRNet优于现有的CsiNet）

A In this paper, we propose a novel neural network called
channel reconstruction network (CRNet) based on multi-
resolution architecture. We also introduce an advanced training
scheme that fit the CSI feedback task better to boost the
performance. Simulation results show that with the same
computational complexity, the proposed CRNet greatly out-
performs the CsiNet especially under the advanced scheme.
The main contribution of this paper is listed below.neural network
named “CsiNet” is designed in [12] and its overwhelming
superiority against traditional CS methods is given. After that,
some researchers expand the original scene. Correlation of
CSI in different time slots is used in [13] and correlation
between the uplink and downlink CSI is utilized in [14], [15].
These can be seen as new sub-scenes with extra conditions or
assumptions.（文献[12]设计了一个名为“CsiNet”的神经网络，并给出了它相对于传统CS方法的压倒性优势。之后，一些研究人员扩展了原始场景。在[13]中使用了不同时隙中的CSI的相关性，在[14]，[15]中使用了上行链路和下行链路CSI之间的相关性。这些可以看作是带有额外条件或假设的新子场景。[12] C. Wen, W. Shih and S. Jin, “Deep Learning for Massive MIMO CSI
Feedback,” IEEE Wireless Communications Letters, vol. 7, no. 5, pp.
748-751, Oct. 2018.[13] T. Wang, C. Wen, S. Jin and G. Y . Li, “Deep Learning-Based CSI
Feedback Approach for Time-V arying Massive MIMO Channels,” IEEE
Wireless Communications Letters, vol. 8, no. 2, pp. 416-419, April 2019.
[14] Y . Yang, F. Gao, G. Y . Li and M. Jian, “Deep Learning based Downlink
Channel Prediction for FDD Massive MIMO System,” IEEE Communi-
cations Letters, in press.[15] Z. Liu, L. Zhang and Z. Ding, “Exploiting Bi-Directional Channel
Reciprocity in Deep Learning for Low Rate Massive MIMO CSI
Feedback,” IEEE Wireless Communications Letters, vol. 8, no. 3, pp.
889-892, June 2019.)

The CsiNetPlus proposed in [16] improves the network
performance without e