跳到主要內容

臺灣博碩士論文加值系統

(44.201.97.0) 您好!臺灣時間:2024/04/16 09:33
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

: 
twitterline
研究生:劉志宏
研究生(外文):Chih-Hung Liu
論文名稱:以卷積長暫態記憶神經網路進行調變分類技術研究
論文名稱(外文):Modulation Classification Using Fully Connected Deep Neural Networks with Convolutional Long Short-Term Memory
指導教授:林嘉慶林嘉慶引用關係
指導教授(外文):Jia-Chin Lin
學位類別:碩士
校院名稱:國立中央大學
系所名稱:通訊工程學系在職專班
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2018
畢業學年度:106
語文別:中文
論文頁數:68
中文關鍵詞:類神經網路調變分類
外文關鍵詞:Neural NetworkModulation classification
相關次數:
  • 被引用被引用:0
  • 點閱點閱:172
  • 評分評分:
  • 下載下載:30
  • 收藏至我的研究室書目清單書目收藏:0
在軍事領域中,如何獲取敵方通訊內容一直是各國投入大量人力與心血研究的課題,獲取敵方通訊內容的第一步,便是將截收到之訊號進行調變分類,後續才能進一步地研究如何破譯並取出其內容,然而現今訊號調變方式越來越多元,如何將訊號快速準確地進行分類,儼然成為一個重要的研究項目。
傳統訊號調變分類方式,仰賴人工運用複雜運算進行特徵擷取,再依照不同特徵進行調變分類,本文透過類神經網路特有的自我學習來進行特徵擷取與分類,跳脫自行擷取所需的複雜運算,將時間與精神專注於類神經網路演算法的改良,提升調變分類的準確度。
本文提出以卷積長暫態記憶神經網路進行調變分類技術研究,將現有之神經網路分別從訓練模型與資料集等2個部分進行改良,提出之改良型卷積長暫期記憶神經網路具有強大的抗雜訊能力與細部特徵擷取能力,經過各式不同的測試,此模型整體調變分類成功率可達64.7%,在訊雜比為0dB~20dB的範圍內,調變分類成功率可以達到90.1%,高訊雜比(+18dB)成功率可達90%,有不錯的效果。
Modulation classification is usually the first step of a major communications problem with military applications. We have to know the modulation types before we decode the signals and get the content. As the modulation types increase rapidly, automatic modulation recognition becomes an important topic which is worth researching into.
Traditionally, we use manual feature selection to get the features and do the classification. In this article, we aim to use of DL to learn from data, extract features and classify signals automatically. We will concentrate on modifying the model of DL to improve the classification accuracy.
This paper proposes a research of modulation classification using fully connected dep neural networks with convolutional long short-term memory. We modify the existing model by improving the training model and dataset. The overall classification accuracy of the modified model is 64.7%. In high SNR region(0dB~20dB), the classification accuracy is 90.1%. In high SNR(+18dB), the classification accuracy is 90%
摘要 I
Abstract II
致謝 III
目錄 IV
圖目錄 VI
表目錄 VIII
第一章 緒論 1
1.1 研究背景 1
1.2 研究動機 2
1.3 論文架構 2
第二章 深度學習簡介 4
2.1 感知器 4
2.2 深度神經網路 5
2.3 卷積神經網路 6
2.3.1 卷積層 6
2.3.2 池化層 8
2.3.3 全連接層 8
2.4 循環神經網路 10
2.4.1 循環神經網路介紹 10
2.4.2 長暫態記憶網路介紹 13
2.5 卷積長暫態記憶神經網路 16
第三章 卷積長暫態記憶神經網路之調變分類 19
3.1 卷積長暫態記憶神經網路進行調變分類 19
3.1.1 訓練模型 19
3.1.2 資料集 21
3.2 卷積長暫態記憶神經網路以不同資料表進行分類 22
3.2.1 訓練模型 22
3.2.2 資料集 24
3.3 卷積長暫態記憶神經網路進一步分析 25
3.3.1 訓練模型 25
3.3.2 資料集 26
第四章 改良之卷積長暫態記憶神經網路 29
4.1. 改良之訓練模型 29
4.1.1. 加入多層長暫態記憶層 29
4.1.2. 最佳化卷積層層數 32
4.2. 改良資料集 37
4.2.1. 提升資料量 37
4.2.2. 修改訓練資料 41
4.3. 結果與分析 47
4.3.1. 模擬環境 47
4.3.2. 改良型卷積長暫態記憶神經網路模型 47
4.3.3. 混淆矩陣分析 52
第五章 結論與未來方向 54
參考文獻 56
[1] Ka.Mun.Ho,C.Vaz,D.G.Daut, “Automatic classification of amplitude, frequency, and phase shift keyed signals in the wavelet domain,” IEEE Sarnoff Symposium, pp.1-6,April 2010.
[2] 林聰岷,《使用高階統計法則實現相位鍵移調變訊號分類作業》,碩士論文,國立臺灣大學電信工程學研究所,2012年1月。
[3] Chih-Wei Hsu, Chih-Chung Chang, and Chih-Jen Lin , “A Practical Guide to Support Vector Classification”,pp.16, May 19, 2016
[4] V. Mitra, W. Wang, and H/ Franco, “Deep Convolutional Nets and Robust Features for Reverberation-robust Speech Recognition,” in Proceedings of SLT, 2014.
[5] S. Wang, L. Chen, L. Xu, W. Fan, J. Sun, S. Naoi, “Deep knowledge training and heterogeneous CNN for handwritten Chinese text recognition,” Proc. ICFHR-2016, pp.84-89,2016.
[6] T. J. O’Shea, J. Corgan, “Convolutional radio modulation recognition networks”, CoRR, vol. abs/1602, no. 04105, pp. 1-15, Mar. 2016.
[7] Shengliang Peng, Hanyu Jiang, Huaxia Wang, Hathal Alwageed, and Yu-Dong Yao, "Modulation Classification Using Convolutional Neural Network Based Deep Learning Model",2017
[8] N.E. West, T. O'Shea, "Deep Architectures for Modulation Recognition", Proc. 2017 IEEE Int'l. Symposium Dynamic Spectrum Access Networks (DySPAN), pp. 16, March 2017.
[9] Hinton G E, Osindero S, Teh Y W. ”A fast learning algorithm for deep belief nets”. Neural computation, 2006, 18(7): 1527-1554.
[10] Christopher Olah.,Understanding LSTM Networks, http://colah.github.io/posts/2015-08-Understanding-LSTMs/,accessed 2017
[11] Tara N. Sainath, Oriol Vinyals, Andrew Senior, Has¸im Sak, “CONVOLUTIONAL, LONGSHORT-TERM MEMORY, FULLY CONNECTED DEEP NEURAL NETWORKS”, Google, Inc,2015.
[12] V. Nair, G.E. Hinton, "Rectified Linear Units Improve Restricted Boltzmann Machines", Proc. Int'l Conf. Machine Learning, 2010.
[13] TaraN Sainath ,Ron J. Weiss, Andrew Senior, Kevin W. Wilson, and Oriol Vinyals.” Learning the speech front-end with raw waveform CLDNNs”, Google, Inc,2015.
[14] Xiaoyu Liu, Diyu Yang, and Aly El Gamal, ” Deep Neural Network Architectures for Modulation Classification”,pp.1-5,2017
[15] T. O'Shea, J. Shea, Open Radio Machine Learning Datasets for Open Science. https://www.deepsig.io/datasets, accessed 2017
[16] T. J. O’Shea and N. West, “Radio machine learning dataset generation with gnu radio” in Proceedings of the GNU Radio Conference, vol. 1, no. 1, 2016.
[17] S. Rajendran, W. Meert, D. Giustiniano, V. Lenders, S. Pollin, “Distributed deep learning models for wireless signal classification with low-cost spectrum sensors.”, arXiv:1707.08908,pp.1-13, 2017
[18] M. Kulin, T. Kazaz, I. Moerman, E. d. Poorter, “End-to-end Learning from Spectrum Data: A Deep Learning approach for Wireless Signal Identification in Spectrum Monitoring applications”, arXiv:1712.03987, 2017
[19] D. Kingma, J. Ba, Adam: A method for stochastic optimization, arXiv:1412.6980, pp.1-15,2014
[20] Justin Alexander, ” OFDM Modulation Recognition Using Convolutional Neural Networks”, degree of Master of Engineering, Apr. 26 2017.
[21] Xuming Lin, Ruifang Liu, Wenmei Hu, Yameng Li,” A Deep Convolutional Network Demodulator for Mixed Signals with Different Modulation Types”, pp.893-896,2017
[22] Yun Lin, Ya Tu , Zheng Dou, Zhiqiang Wu, “The Applicaction of Deep Learning in Communication Signal Modulation Recognition” in International Conference on Communications in China, 2017.
[23] Google Research Blog “Announcing Tensorflow 0.8 – now with distributed computing support!”. https://ai.googleblog.com/2016/04/announcing-tensorflow-08-now-with.html, accessed 2017.
[24] S. J. Pan and Q. Yang, “A survey on transfer learning,” Knowledge and Data Engineering, IEEE Transactions on, vol. 22, no. 10, pp. 1345– 1359, 2010.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top