研究生: |
梁松澤 Song-Ze Liang |
---|---|
論文名稱: |
深度學習方法解決脈衝在極化碼上的研究 Deep Learning Method to Solve the Research of Impulse on Polar Code |
指導教授: |
曾德峰
De-Feng Tseng |
口試委員: |
陳永芳
Yung-Fang Chen 曾恕銘 Shu-Ming Tseng 張立中 Li-Jung Jang 曾德峰 De-Feng Tseng |
學位類別: |
碩士 Master |
系所名稱: |
電資學院 - 電機工程系 Department of Electrical Engineering |
論文出版年: | 2022 |
畢業學年度: | 110 |
語文別: | 中文 |
論文頁數: | 92 |
中文關鍵詞: | 錯誤更正碼 、極化碼解碼器 、脈衝雜訊通道 、全連結式神經網路 、微調訓練方法 |
外文關鍵詞: | Error–Correcting Codes, Polar code decoder, Impulse noise channel, Fully-connected neural network, Fine-Tuning training method |
相關次數: | 點閱:415 下載:10 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
本文主要目標在探討深度學習極化碼(Polar Code)解碼器是否可以在脈衝雜訊(Impulse Noise)干擾下,能透過全連接式神經網路的學習,讓接收器可以有更好的更正錯誤能力。
對於有線通訊或是無線通訊,在傳輸的過程中,脈衝雜訊的出現與干擾是無可避免的,脈衝雜訊瞬間強力的干擾不但會破壞傳送中的訊號,更會使得接收端在收到傳輸訊號後解讀錯誤的機率大增。
本文將介紹雜訊通道模型有兩種 : 第一種是常見的可加性高斯白雜訊(Additive White Gaussian Noise, AWGN)通道模型,第二種是白努利-高斯脈衝雜訊(Bernoulli-Gaussian impulse, BG-impulse)通道模型。
本文將使用全連接式深度神經網路的架構,來模擬訊號經過極化碼(Polar Code)編碼後,通過雜訊通道干擾後解碼的性能。以AWGN所訓練出來的模型作為比較,再介紹微調(Fine-Tuning)的訓練方法,來訓練脈衝雜訊模型,提升受脈衝干擾的抵抗性。
The main goal of this paper is to investigate whether a deep learning Polar Code decoder can provide better error correction capability to the receiver by learning from a fully-connected neural network under the interference of Impulse Noise.
For wired or wireless communications, the appearance of impulse noise and interference during transmission is inevitable. The transient and powerful interference of impulse noise not only destroys the transmitted signal, but also increases the chance of misinterpretation at the receiving end after receiving the transmitted signal.
In this paper, we introduce two types of noise channel models: the first is the common Additive White Gaussian Noise (AWGN) channel model, and the second is the Bernoulli-Gaussian impulse(BG-impulse) channel model.
In this paper, we use a fully-connected deep neural network (FCDNNs) architecture to simulate the performance of decoding signals after they have been encoded by polarized codes and interfered by noise channels. The model trained by AWGN is used as a comparison, and Fine-Tuning is introduced to train the impulse noise model to improve the resistance to pulse interference.
[1] C. E. Shannon, "A mathematical theory of communication," The Bell System Technical Journal, vol. 27, no. 4, pp. 623-656, 1948.
[2] S. L. a. D. J. Costello, “Error Control Coding: Fundamentals and Applications.” Pearson-Prentice Hall, 2004.
[3] E. Arikan, "Channel Polarization: A Method for Constructing Capacity-Achieving Codes for Symmetric Binary-Input Memoryless Channels," IEEE Transactions on Information Theory, vol. 55, no. 7, pp. 3051-3073, 2009.
[4] Tobias Gruber, Sebastian Cammerer, Jakob Hoydis, Stephan ten Brink, “On Deep Learning-Based Channel Decoding,” arXiv, Submitted on 26 Jan 2017
[5] X.-A. Wang and S. B. Wicker, “An artificial neural net Viterbi decoder,”
IEEE Trans. Commun., vol. 44, no. 2, pp. 165–171, Feb. 1996.
[6] E. Arikan, "Channel polarization: A method for constructing capacity-achieving codes," in 2008 IEEE International Symposium on Information Theory, 2008, pp. 1173-1177.
[7] E. Arikan, "Channel combining and splitting for cutoff rate improvement," IEEE Transactions on Information Theory, vol. 52, no. 2, pp. 628-639, 2006.
[8] Harish Vangala, Emanuele Viterbo, Yi Hong, “A Comparative Study of Polar Code Constructions for the AWGN Channel,” arXiv, Submitted on 11 Jan 2015
[9] H. Li and J. Yuan, "A practical construction method for polar codes in AWGN channels," in IEEE 2013 Tencon - Spring, 2013, pp. 223-226.
[10] Ido Tal, Alexander Vardy, “List Decoding of Polar Codes,” arXiv, Submitted on 31 May 2012
[11] Kai Chen, Kai Niu, Jia-Ru Lin, “Improved Successive Cancellation Decoding of Polar Codes,” arXiv, Submitted on 17 Aug 2012 (v1), last revised 17 Jan 2013 (this version, v2)
[12] Bo Yuan and Keshab K. Parhi, “Successive Cancellation List Polar Decoder using Log-likelihood Ratios,” IEEE, 2014 48th Asilomar Conference on Signals, Systems and Computers
[13] Seyyed Ali Hashemi, Carlo Condo, Warren J. Gross, “Simplified Successive-Cancellation List Decoding of Polar Codes,” 2016 IEEE International Symposium on Information Theory
[14] Alexios Balatsoukas-Stimming, Student Member, “LLR-Based Successive Cancellation List Decoding of Polar Codes,” IEEE Transactions on Signal Processing, 2015, 63(19):5165-5179.
[15] 林尹泰, "極化碼基於不同解碼方式之效能分析比較," 國立台灣科技大學電機工程研究所碩士論文, 民國一百零八年七月.
[16] Lei Jimmy Ba, Rich Caruana, “Do Deep Nets Really Need to be Deep?,” Submitted on 21 Dec 2013 (v1), last revised 11 Oct 2014 (this version, v7)
[17] Hyeji Kim, Yihan Jiang, Ranvir Rana, Sreeram Kannan, Sewoong Oh, Pramod Viswanath, “Communication Algorithms via Deep Learning,” University of Washington, May 2018.
[18] 林哲賢, “OFDM系統在多路徑衰減下探討神經網路接收器遭遇脈衝干擾之性能,” 國立臺灣科技大學電機工程研究所碩士論文, 民國一百零九年十二月