大規模多輸入多輸出 (Massive Multi-Input-Multi-Output，簡稱 Massive MIMO)系統在分頻雙工模式下，需透過回傳鏈路將下行鏈路的通道狀態資訊 (Channel State Information，簡稱 CSI) 發送至基地台端，如此才能使得Massive MIMO得以發揮其潛在增益；但是，由於回傳鏈路頻寬的限制，因此若要將完整的 CSI 回傳至基地台端，幾乎是無法達成的任務。本論文中，我們利用深度學習的技術來建立一新型的 CSI 壓縮與重建的類神經網路，稱作 CsiNet (CSI Sensing and Recovery Network)。透過大量的訓練資料，CsiNet 可以學習如何有效地利用通道架構將 CSI 轉換成碼字以及由碼字到 CSI 的逆轉換。透過模擬我們將CsiNet 的效能與現存壓縮感知 (Compressive Sensing，簡稱 CS) 的演算法做比較，說明 CsiNet 的重建結果明顯優於現存CS的效能。不僅如此，CS 演算法在過低的壓縮率時是無法還原 CSI，然而 CsiNet 仍保持著一定的波束成形增益。

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 thesis, 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.

論文審定書 i
論文公開授權書 ii
誌謝 iii
摘要 iv
Abstract v
目錄 vi
圖次 viii
表次 ix
第一章 緒論 1
1.1 研究背景與動機 1
1.2 論文安排 3
第二章 相關背景 4
2.1 正交分頻多工簡介 4
2.2 多輸入多輸出技術 4
2.3 系統模型 5
2.4 壓縮感知 7
第三章 利用深度學習的回傳 8
3.1 深度學習 8
3.1.1 卷積層 (Convolution Layer) 9
3.1.2 激勵函數 (Activation Function) 10
3.1.3 池化層 (Pooling Layer) 11
3.2 深度學習的回傳機制 12
3.2.1 編碼端 12
3.2.2 解碼端 12
3.3 通道狀態資訊回傳網路結構 (CsiNet) 13
3.3.1 RefineNet 15
3.3.2 批標準化 (Batch normalize) 15
3.4 通道狀態資訊回傳全卷基層網路結構 (ConvCsiNet) 16
3.4.1 雙線性升採樣 (Bilinear Up-Sampling) 19
3.5 部分通道資訊輔助之深度學習回傳方法 20
3.5.1 長短期記憶 (Long Short-Term Memory) 22
第四章 模擬結果與分析 24
4.1 資料生成 24
4.2 模擬結果 26
4.2.1 CS 演算法與 CsiNet-based 之效能與時間複雜度比較 27
4.2.2 不同傳送天線數之效能比較 33
4.2.3 不同域的 CsiNet-based 結果比較 34
4.2.4 有無部份通道資訊輔助的效能比較 36
第五章 結論 37
參考文獻 38

[1] T. L. Marzetta, “Noncooperative cellular wireless with unlimited numbers of base station antennas,” IEEE Trans. Wireless Commun., vol. 9, no. 11, pp. 3590– 3600, Nov. 2010.
[2] J. Zhang, C. K. Wen, S. Jin, X. Gao, and K. K. Wong, “On capacity of large-scale MIMO multiple access channels with distributed sets of correlated antennas,” IEEE J. Sel. Areas Commun., vol. 31, no. 2, pp. 133– 148, Feb. 2013.
[3] P. H. Kuo, H. Kung, and P. A. Ting, “Compressive sensing based channel feedback protocols for spatially-correlated massive antenna arrays,” in Proc. IEEE WCNC, Shanghai, China, Apr. 2012, pp. 492– 497.
[4] X. Rao and V. K. Lau, “Distributed compressive CSIT estimation and feedback for FDD multi-user massive MIMO systems,” IEEE Trans. Signal Process., vol. 62, no. 12, pp. 3261– 3271, Jun. 2014.
[5] L. Liu, C. Oestges, J. Poutanen, K. Haneda, P. Vainikainen, F. Quitin, F. Tufvesson, and P. D. Doncker, “The COST 2100 MIMO channel model,” IEEE Wireless Commun., vol. 19, no. 6, pp. 92– 99, Dec. 2012.
[6] I. Daubechies, M. Defrise, and C. D. Mol, “An iterative thresholding algorithm for linear inverse problems with a sparsity constraint,” Comm. Pure and Applied Math., vol. 75, pp. 1412– 1457, 2004.
[7] D. L. Donoho, A. Maleki, and A. Montanari, “Message passing algorithms for compressed sensing,” Proc. Natl. Acad. Sci., vol. 106, no. 45, pp. 18 914– 18 919, 2009.
[8] C. Li, W. Yin, and Y. Zhang, “User’s guide for tval3: Tv minimization by augmented lagrangian and alternating direction algorithms,” CAAM report, vol. 20, pp. 46– 47, 2009.
[9] C. A. Metzler, A. Maleki, and R. G. Baraniuk, “From denoising to compressed sensing,” IEEE Trans. Inf. Theory, vol. 62, no. 9, pp. 5117– 5144, 2016.
[10] Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner, “Gradient-based learning applied to document recognition,” Proceedings of the IEEE, vol. 86, no. 11, pp. 2278–2324, 1998.
[11] I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. MIT Press.
[12] Timothy J. O'Shea, Tugba Erpek, and T. Charles Clancy, “Deep learning based MIMO communications,” preprint, 2017. [Online]. Available: https://arxiv.org/abs/1707.07980.
[13] S. Lohit, K. Kulkarni, R. Kerviche, P. Turaga, and A. Ashok, “Convolutional neural networks for non-iterative reconstruction of compressively sensed images,” preprint, 2017. [Online]. Available: http://arxiv.org/abs/1708.04669.
[14] A. Mousavi, G. Dasarathy, and R. G. Baraniuk, “DeepCodec: Adaptive sensing and recovery via deep convolutional neural networks,” preprint, 2017. [Online]. Available: http://arxiv.org/abs/1707.03386.
[15] H. Yao, F. Dai, D. Zhang, Y. Ma, S. Zhang, and Y. Zhang, “DR2-Net: Deep residual reconstruction network for image compressive sensing,” preprint, 2017. [Online]. Available: http://arxiv.org/abs/1702.05743.
[16] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in CVPR, 2016, pp. 770– 778.
[17] A. Krizhevsky, I. Sutskever, and G. Hinton, “Imagenet classification with deep convolutional neural networks,” in NIPS, 2012, pp. 1097– 1105.
[18] K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” preprint, 2014. [Online]. Available: https://arxiv.org/abs/1409.1556.
[19] M. D. Zeiler and R. Fergus, “Visualizing and understanding convolutional networks,” in ECCV, 2014, pp. 818– 833.
[20] J. Long, E. Shelhamer, and T. Darrell, “Fully convolutional networks for semantic segmentation,” in CVPR, 2015, pp. 3431– 3440.
[21] W. Zaremba, I. Sutskever, and O. Vinyals, “Recurrent neural network regularization,” preprint, 2014. [Online]. Available: http://arxiv.org/abs/1409.2329.
[22] S. Xingjian, Z. Chen, H. Wang, D. Yeung, W. Wong, and W. Woo, “Convolutional LSTM network: A machine learning approach for precipitation nowcasting,” in NIPS, 2015, pp. 802– 810.
[23] C. K. Wen, W. T. Shih, and S. Jin, “Deep learning for massive MIMO CSI feedback,” IEEE Wireless Commun. Lett., 2018. DOI: 10.1109/LWC.2018.2818160