基於深度學習辨識跳頻信號之研究

以作者查詢圖書館館藏

、以作者查詢臺灣博碩士

、以作者查詢全國書目

、勘誤回報

、線上人數：18

、訪客IP：18.219.13.74

姓名

饒廣衡(Kuang-Heng Jao) 查詢紙本館藏

畢業系所

資訊管理學系在職專班

論文名稱

基於深度學習辨識跳頻信號之研究

相關論文

★ 台灣50走勢分析：以多重長短期記憶模型架構為基礎之預測	★ 以多重遞迴歸神經網路模型為基礎之黃金價格預測分析
★ 增量學習用於工業4.0瑕疵檢測	★ 遞回歸神經網路於電腦零組件銷售價格預測之研究
★ 長短期記憶神經網路於釣魚網站預測之研究	★ Opinion Leader Discovery in Dynamic Social Networks
★ 深度學習模型於工業4.0之機台虛擬量測應用	★ A Novel NMF-Based Movie Recommendation with Time Decay
★ 以類別為基礎sequence-to-sequence模型之POI旅遊行程推薦	★ A DQN-Based Reinforcement Learning Model for Neural Network Architecture Search
★ Neural Network Architecture Optimization Based on Virtual Reward Reinforcement Learning	★ 生成式對抗網路架構搜尋
★ 以漸進式基因演算法實現神經網路架構搜尋最佳化	★ Enhanced Model Agnostic Meta Learning with Meta Gradient Memory
★ 遞迴類神經網路結合先期工業廢水指標之股價預測研究	★ A Novel Reinforcement Learning Model for Intelligent Investigation on Supply Chain Market

檔案

[Endnote RIS 格式]

[Bibtex 格式]

[相關文章]

[文章引用]

[完整記錄]

[館藏目錄]

至系統瀏覽論文 ( 永不開放)

摘要(中)

無線電通信仰賴在各個不同頻譜上傳遞各項訊息，為了避免遭受通信干擾狀況，研發出新式跳頻通信，跳頻通信因為在每個頻率上停留的時間極短，廣泛的運用在各式各樣的通信頻段（如WiFi無線網際網路、藍芽通信及無人機操控等設備）上，其主要原因是跳頻通信較不易受到外在的影響，且不易被偵測到，以往傳統定頻無線電藉由使用頻率、聲音特徵、出網次數等數據來分析、判斷，近年來深度學習在影像辨識技術進步飛快，本研究希望開發出一個非線性深度學習模型，將即時頻譜轉換為頻譜圖透過快速影像辨識方式，快速完成識別後，應可發揮實際效用。
本研究商借國家中山科學研究院儀器設備，藉產生型態近似相同而速率不同之跳頻信號，透過深度學習模型實施頻譜圖影像辨識，實驗結果確實可達到高準確率，如能持續增加不同類別信號及資料實施訓練，當能發揮更佳效益。

摘要(英)

Radio communication relies on transmitting various information on different frequency spectrums. In order to avoid communication interference, a new type of frequency hopping communication has been developed. Because frequency hopping communication stays on each frequency for a very short time, it is widely used in a variety of The main reason is that frequency hopping communication is less susceptible to external influences and is not easy to be detected. In the past, traditional fixed-frequency radios By using data such as frequency, sound characteristics, and network access times to analyze and judge, in recent years, deep learning has made rapid progress in image recognition technology. Identification method, after the identification is completed quickly, it will be able to play a practical role.
In this research, the instruments and equipment of the National Chung-Shan Institute of Science and Technology(NCSIST) are used to generate frequency hopping signals with approximately the same type but different rates. The deep learning model is used to implement spectrogram image recognition. The experimental results can indeed achieve high accuracy. Class signals and data implementation training should be able to play a substantial effect.

關鍵字(中)

★ 跳頻
★ 頻譜圖
★ 深度學習
★ 卷積式神經網路

關鍵字(英)

論文目次

目錄
第一章緒論 1
1.1 研究背景 1
1.2 研究動機與目的 2
1.3 研究貢獻 3
1.4 論文架構 4
第二章文獻及背景 5
2.1. 跳頻（FREQUENCY HOPPING）信號理論與時頻圖建立 6
2.2. 跳頻（FREQUENCY HOPPING）信號辨識 8
2.3. 機器學習 10
2.4. 機器學習於跳頻辨識相關研究 14
2.5. 相關研究分析比較 20
第三章研究方法 22
3.1. 研究架構 22
3.2. 跳頻（FREQUENCY HOPPING）信號資料集 23
3.3. 資料前處理 24
3.4. 模型建立與訓練 26
第四章模型實驗 30
4.1 實驗環境 30
4.2 實驗資料集 30
4.3 模型評估指標 31
4.4 模型效能之比較 32
4.5 調整參數模型比較之研究 33
4.6 實驗總結 36
第五章結論 38
5.1 研究總結 38
5.2 研究限制 38
5.3 未來研究方向 38
參考文獻 40

參考文獻

[1] Nandi, A. K., & Azzouz, E. E. (1998). Algorithms for automatic modulation recognition of communication signals. IEEE Transactions on communications, 46(4), 431-436.
[2] 林聰岷. (2013). 使用高階統計法則實現相位鍵移調變訊號分類作業.
[3] Dobre, O. A., Bar-Ness, Y., & Su, W. (2003, October). Higher-order cyclic cumulants for high order modulation classification. In IEEE Military Communications Conference, 2003. MILCOM 2003. (Vol. 1, pp. 112-117). IEEE.
[4] Lee, K. G., & Oh, S. J. (2020). Detection of frequency-hopping signals with deep learning. IEEE Communications Letters, 24(5), 1042-1046.
[5] Bracewell, R. N., & Bracewell, R. N. (1986). The Fourier transform and its applications (Vol. 31999, pp. 267-272). New York: McGraw-hill.
[6] 趙俊, 張朝陽, 賴利峰, & 曹千芊. (2003). 一種基於時頻分析的跳頻信號參數盲估計方法. 電路與系統學報, 8(3), 46-50.
[7] 郝威, & 楊露菁. (2004). 跳頻技術的發展及其干擾對策. 艦船電子對抗, 27(4), 7-12.
[8] 跳頻通信:跳頻是最常用的擴頻方式之一，其工作原理是指收發雙方傳輸信號的載 -百科知識中文網(N.d.). https://www.easyatm.com.tw/wiki/%E8%B7%B3%E9%A0%BB%E9%80%9A%E4%BF%A1
[9] 侯范, 姚志成, 楊劍, 李昱婷, & 王自維. (2022). 一種基於K-means 聚類的跳頻信號快速檢測方法. Telecommunication Engineering, 62(2).
[10] Griffin, D., & Lim, J. (1984). Signal estimation from modified short-time Fourier transform. IEEE Transactions on acoustics, speech, and signal processing, 32(2), 236-243.
[11] 劉佳敏, 趙知勁, 曹越飛, 葉學義, & 王李軍. (2021). 基於時頻分析的多跳頻信號盲檢測. Journal of Signal Processing, 37(5).
[12] 李紅光, 郭英, 齊子森, & 蘇令華. (2020). 複雜電磁環境下多跳頻信號盲檢測. 華中科技學報: 自然科學版, 48(7), 13-19.
[13] Recognition of Overlapped Frequency Hopping Signals Based on Fully Convolutional Networks.(Pengcheng Liu et al.2021 )
[14] Book_李宏毅老師機器學習課程筆記 – HackMD (N.d.). https://hackmd.io/@shaoeChen/B1CoXxvmm/https%3A%2F%2Fhackmd.io%2Fs%2FHyKhr5sRz
[15] 國立台灣大學計算機及資訊網路中心電子報(N.d.). https://www.cc.ntu.edu.tw/chinese/epaper/0038/20160920_3805.html
[16] AI - Ch16 機器學習(4), 類神經網路 Neural network　(N.d.). https://mropengate.blogspot.com/2015/06/ch15-4-neural-network.html
[17] Liu, P., Han, Z., Shi, Z., & Liu, M. (2021, June). Recognition of overlapped frequency hopping signals based on fully convolutional networks. In 2021 28th International Conference on Telecommunications (ICT) (pp. 1-5). IEEE.
[18] 呂國裴, & 謝躍雷. (2020). 基於深度學習的跳頻信號識別. Telecommunication Engineering, 60(10).
[19] Lee, K. G., & Oh, S. J. (2020). Detection of frequency-hopping signals with deep learning. IEEE Communications Letters, 24(5), 1042-1046.
[20] Li, Z., Liu, R., Lin, X., & Shi, H. (2018, December). Detection of frequency-hopping signals based on deep neural networks. In 2018 IEEE 3rd International Conference on Communication and Information Systems (ICCIS) (pp. 49-52). IEEE.
[21] Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L. C. (2018). Mobilenetv2: Inverted residuals and linear bottlenecks. In Proceedings of the IEEE conference on computer vision and pattern recognition (pp. 4510-4520).
[22] 初探卷積神經網路 (N.d.). https://chtseng.wordpress.com/2017/09/12/%E5%88%9D%E6%8E%A2%E5%8D%B7%E7%A9%8D%E7%A5%9E%E7%B6%93%E7%B6%B2%E8%B7%AF/
[23] Ithome網站資訊 (N.d.).
https://ithelp.ithome.com.tw/articles/10263847

指導教授

陳以錚

審核日期

2022-7-12

推文