基於深度學習之多用戶上行低軌道衛星通訊定時提前預測

簡易檢索 / 詳目顯示

回結果列表

研究生：	呂亭儀 Lu, Ting-Yi
論文名稱：	基於深度學習之多用戶上行低軌道衛星通訊定時提前預測 Deep Learning Based Timing Advance Prediction for Multi-User Uplink LEO Satellite Communication
指導教授：	吳仁銘 Wu, Jen-Ming
口試委員:	張佑榕 Chang, Ronald Y. 鍾偉和 Chung, Wei-Ho
學位類別：	碩士 Master
系所名稱：	電機資訊學院 - 通訊工程研究所 Communications Engineering
論文出版年：	2024
畢業學年度：	112
語文別：	英文
論文頁數：	59
中文關鍵詞：	低軌道衛星通訊、定時提前、上行鏈路同步、神經網路結構、深度學習
外文關鍵詞：	LEO SatCom, Timing advance, Uplink timing synchronization, Neural network architecture, Deep learning
相關次數：	點閱：3 下載：0
分享至:	分享至facebook 分享至twitter

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

本文提出了一種利用深度學習預測低地球軌道（LEO）衛星通訊系統上行鏈路(uplink)定時提前（TA）的方法，以實現正交分頻多重進接（OFDMA）系統上行鏈路的時間同步。

上行鏈路時間同步對地面網絡（TN）和非地面網絡（NTN）是一個極為重要的問題。缺乏同步可能導致多個用戶之間引發多重進接干擾（MAI）。為確保5G NR中細胞網路上行鏈路的正交性，要求同一子幀（subframe）內不同用戶發送的信號在到達基地台（BS）時保持時間上的對齊。

在LEO衛星通信系統中，由於衛星以約7.6 km/s的高速度移動，導致星地鏈路（satellite-to-ground link）中的傳播延遲呈現動態變化。這樣的變化會造成衛星發送的時間提前命令（Timing Advance Command）在用戶終端（UT）接收到時可能已經過時。因此，現有地面通信估計定時提前（TA）的方法在動態傳播環境中不再適用。

在本論文中，我們提出了基於深度學習的衛星定時提前預測器。該預測器首先進行離線學習以進行訓練，然後根據輸入資料即時回饋相對應的輸出結果。數值結果顯示，我們提出的基於深度學習的預測器能夠有效預測每個UT未來時間與衛星之間傳播延遲的變化。讓UT判斷需要提前多少時間傳送信號到衛星端，實現上行鏈路的同步。通過仿真驗證，我們的方法效能良好，並且適用於沒有衛星導航系統(GNSS)的用戶終端。該方法成功解決了在LEO衛星通訊系統中用戶終端上行鏈路同步的問題，進而提高了OFDMA系統的性能。

This paper proposes the utilization of deep learning to estimate the Time Advance (TA) of Low Earth Orbit (LEO) satellite communication systems. This estimation is crucial for enabling the Orthogonal Frequency Division Multiple Access (OFDMA) system to achieve uplink synchronization.

Uplink time synchronization is a critical concern for terrestrial networks (TN) and non-terrestrial network (NTN). Without proper synchronization, interference among multiple users and downlink imbalances can lead to Multiple Access Interference (MAI). To maintain orthogonality in the uplink within the cell, 5G NR mandates that signals from different users in the same subframe align at the time they reach the base station (BS).

In the LEO satellite communication system, the high speed of satellite movement (approximately 7.6 km/s) introduces dynamic changes in the propagation delay within the satellite-to-ground link. This dynamic nature causes the Timing Advance Command to become outdated by the time it is received by the User Terminal (UT). Consequently, existing TA estimation methods designed for terrestrial communications are not suitable for such dynamic propagation environments. Additionally, due to the satellite's distance of about 500-2000 kilometers from the ground-based UT, the propagation delay in satellite communication significantly exceeds that in ground communication. Consequently, existing preamble sequences are insufficiently long to cover the larger propagation delays in satellite-to-ground links.

This paper explores several deep learning neural network algorithms to design the proposed Satellite Timing Advance Predictor. The predictor undergoes initial training through offline learning, and subsequently, the corresponding output results are fed back online based on the input data. Numerical results demonstrate that the proposed deep learning-based predictor effectively anticipates changes in propagation delay between each UT and the satellite in the future. This enables the UT to determine the necessary time advance for transmitting signals to the satellite, achieving uplink synchronization. Simulations confirm that the proposed method exhibits excellent performance, a low error rate, and is suitable for low signal-to-noise ratio environments. It successfully addresses the challenge of uplink synchronization for earth UTs, thereby enhancing the performance of OFDMA systems.

Abstract
Content
   INTRODUCTION
1    Foreword ---------------------------------------------------------------------------- 1
2    Research Motivation and Objective --------------------------------------------- 3
3    Related Work ----------------------------------------------------------------------- 3
3.1    Utilization of UT Location Information ---------------------------------- 4
3.2    Only Relying on Preamble ------------------------------------------------- 4
4    Proposed Method ------------------------------------------------------------------ 5
5    Contribution and Achievement -------------------------------------------------- 6
6    Thesis Organization --------------------------------------------------------------- 6
   BACKGROUND
1    Non-Terrestrial Networks  ------------------------------------------------------ 7
1.1    The Advantages of NTN ---------------------------------------------------- 8
1.2    Use Case of NTN ------------------------------------------------------------ 9
2    Classification and Characteristics of Satellite Communication ------------- 9
2.1    Different Orbits and Real-World Uses ------------------------------------ 9
2.2    Different Types Of Satellites By Their Functions    --------------------- 11
3    Timing Advanced Estimation    --------------------------------------------------- 12
3.1    Timing Advance Estimation in 5G NR    --------------------------------- 12
3.2    Timing Advance Estimation in SatCom    --------------------------------- 13
4    Satellite Architectures ------------------------------------------------------------ 14
4.1    Regenerative Satellite Architectures --------------------------------------14
4.2    NTN with Transparent Satellite Architectures -------------------------- 15
5    Satellite Visibility Coverage ----------------------------------------------------- 17
5.1    The Speed of The Satellite ------------------------------------------------- 18
5.2    Satellite Visibility Window Duration ------------------------------------- 20
6    Timing Advance In Satellite Communication     -------------------------------- 21
6.1    The Influence of Different Orbital Altitude ----------------------------- 21
6.2    Multi-User Uplink in LEO SatCom -------------------------------------- 22
   DEEP LEARNING BASED TIMING ADVANCE PREDICTOR DESIGN FOR LEO SATELLITE COMMUNICATION
1    Transceiver System Model ------------------------------------------------------- 25
1.1    Signal Model ----------------------------------------------------------------- 26
1.2    Channel Model in Satellite Communication ----------------------------- 28
2    Propagation Delay Establishment ----------------------------------------------- 29
3    DL-Based Propagation Timing Advance Predictor With Offline Training-------- 31
3.1    Training Dataset ------------------------------------------------------------- 32
3.2    Test Dataset ----------------------------------------------------------------- 34
3.3    The Proposed LSTM Architecture    --------------------------------------- 35
4    Model Training Setting ---------------------------------------------------------- 40
5    Timing Advance Prediction for Uplink Timing Synchronization ---------- 41
   NUMERICAL AND SIMULATION RESULTS
1    Simulation Parameters ----------------------------------------------------------- 44
2    Analyzing Time-Varying Propagation Delay Prediction -------------------- 45
3    Compare Performance ----------------------------------------------------------- 51
4    Bit Error Rate Performance ----------------------------------------------------- 54
   CONCLUSIONS ----------------------------------------------------------------------- 56
Bibliography ---------------------------------------------------------------------------------- 57

                                

[1] 3GPP, “Study on new radio (nr) to support non-terrestrial networks,” Technical report (TR) 38.811, 3rd Generation Partnership Project (3GPP), September 2020. (Release 15).
[2] O. Kodheli, A. Guidotti, and A. Vanelli-Coralli, “Integration of satellites in 5g through leo constellations,” in GLOBECOM 2017-2017 IEEE Global Communications Confer- ence, pp. 1–6, IEEE, 2017.
[3] W. Liu, X. Hou, J. Wang, L. Chen, and S. Yoshioka, “Uplink time synchronization method and procedure in release-17 nr ntn,” in 2022 IEEE 95th Vehicular Technology Conference: (VTC2022-Spring), pp. 1–5, 2022.
[4] 3GPP, “On ntn synchronization, random access, and timing advance,” in 3GPP TSG RAN WG1 Meeting 98bis, Chongqing, China,, no. R1-1910982, Oct. 2019.
[5] Y. Zhang, Y. Wu, A. Liu, X. Xia, T. Pan, and X. Liu, “Deep learning-based chan- nel prediction for leo satellite massive mimo communication system,” IEEE Wireless Communications Letters, vol. 10, no. 8, pp. 1835–1839, 2021.
[6] 3GPP, “3rd generation partnership project; technical specification group services and system aspects; release 17 description; summary of rel-17 work items (release 17),” Tech. Rep. TR 21.917, 3GPP, Sep. 2022.
[7] W. Wang, T. Chen, R. Ding, G. Seco-Granados, L. You, and X. Gao, “Location-based timing advance estimation for 5g integrated leo satellite communications,” IEEE Trans- actions on Vehicular Technology, vol. 70, no. 6, pp. 6002–6017, 2021.
[8] H. Yizhou, C. Gaofeng, L. Pengxu, C. Ruijun, and W. Weidong, “Timing advanced estimation algorithm of low complexity based on dft spectrum analysis for satellite system,” China Communications, vol. 12, no. 4, pp. 140–150, 2015.
[9] G. Cui, Y. He, P. Li, and W. Wang, “Enhanced timing advanced estimation with symmetric zadoff-chu sequences for satellite systems,” IEEE Communications Letters, vol. 19, no. 5, pp. 747–750, 2015.
[10] L. Zhen, H. Qin, B. Song, R. Ding, X. Du, and M. Guizani, “Random access preamble design and detection for mobile satellite communication systems,” IEEE Journal on Selected Areas in Communications, vol. 36, no. 2, pp. 280–291, 2018.
[11] I. Goodfellow, Y. Bengio, and A. Courville, Deep learning. MIT press, 2016.
[12] F. Rinaldi, H.-L. Maattanen, J. Torsner, S. Pizzi, S. Andreev, A. Iera, Y. Koucheryavy, and G. Araniti, “Non-terrestrial networks in 5g & beyond: A survey,” IEEE Access, vol. 8, pp. 165178–165200, 2020.
[13] 3GPP, “Solutions for nr to support non-terrestrial networks (ntn),” Technical report (TR) 38.821, 3rd Generation Partnership Project (3GPP), May 2021. (Release 16).
[14] L. Ma, C. Park, X. Wang, P. Gaal, and A. R. Alvarino, “Uplink time synchronization for non-terrestrial networks,” IEEE Communications Magazine, vol. 61, no. 7, pp. 114–118, 2023.
[15] I. Ali, N. Al-Dhahir, and J. Hershey, “Doppler characterization for leo satellites,” IEEE Transactions on Communications, vol. 46, no. 3, pp. 309–313, 1998.
[16] W. Wang, Y. Tong, L. Li, A.-A. Lu, L. You, and X. Gao, “Near optimal timing and frequency offset estimation for 5g integrated leo satellite communication system,” IEEE Access, vol. 7, pp. 113298–113310, 2019.

簡易檢索 / 詳目顯示

相關論文