請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/98091完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 陳昭宏 | zh_TW |
| dc.contributor.advisor | Jau-Horng Chen | en |
| dc.contributor.author | 王盛昱 | zh_TW |
| dc.contributor.author | Sheng-Yu Wang | en |
| dc.date.accessioned | 2025-07-24T16:09:34Z | - |
| dc.date.available | 2025-07-25 | - |
| dc.date.copyright | 2025-07-24 | - |
| dc.date.issued | 2025 | - |
| dc.date.submitted | 2025-07-10 | - |
| dc.identifier.citation | Texas Instruments, “DLP9000X digital micromirror device data sheet,” tech. rep.,Texas Instruments, 2016. https://www.ti.com/lit/ds/symlink/dlp9000x.pdf.
ARMLtd., “AMBA axi protocol specification, revision 4.0,” tech. rep., ARM Ltd., 2010. IEEE 802.3 Working Group, “IEEE Standard for Ethernet Amendment 4: Media Access Control Parameters, Physical Layers, and Management Parameters for 10 Gb/s Operation,” 2002. A. Forencich, “Verilog-Ethernet: A Configurable FPGA Ethernet Core,” GitHub Repository, 2023. https://github.com/alexforencich/verilog-ethernet. ZEMAX Support, “How to model DMD MEMS in Optic Studio. ”https://support.zemax.com/hc/en-us/articles/4414655994899-How-to-model-DMD-MEMS-in-OpticStudio, 2022. J. Postel, “User Datagram Protocol,” 1980. J. Postel, “Transmission Control Protocol,” 1981. T. Nakamura, K. Hashimoto, and S. Yamada, “Limago: A Lightweight FPGA-based 10GbE NIC for Low-Latency Network Processing,” in 2021 IEEE International Conference on Field-Programmable Logic and Applications (FPL), pp. 134–141, 2021. A. Forencich, M. Lavrinc, and E. Petersen, “Corundum: An Open-Source 100GbE FPGA-Based SmartNIC,” ACM Transactions on Reconfigurable Technology and Systems, vol. 15, no. 3, pp. 1–15, 2022. M. Blott, K. Vissers, J. Zambreno, and I. Zecena, “SNAP: An Open FPGA-based Platform for PCIe Networking and Storage Acceleration,” in 2018 IEEE Symposium on High-Performance Interconnects (HOTI), pp. 24–31, 2018. AMD, “7 Series Product Selection Guide.” https://docs.amd.com/v/u/en-US/7-series-product-selection-guide. [Online; accessed 20-Feb-2025]. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/98091 | - |
| dc.description.abstract | 數位微鏡裝置(Digital Micromirror Device, DMD)廣泛應用於投影顯示、光學計算等領域,對即時性與高吞吐量的數據傳輸需求極為嚴苛。在此論文中,數據由電腦端的 Socket 傳送至 FPGA(Field Programmable Gate Array),再經由 FPGA 進行數據處理後,透過 FMC(Field-Programmable Mezzanine Card)接口傳送至 DMD 顯示。為了高效率的數據傳輸,選擇適合的通訊協定至關重要。乙太網協議(Ethernet Protocol)中的 UDP(User Datagram Protocol)與 TCP(Transmission Control Protocol)是常用的解決方案,兩者在傳輸效率與錯誤控制能力上各有優劣。
在乙太網協議中,UDP雖然能提供高吞吐量,但缺乏錯誤控制機制,容易導致數據遺失或錯誤,進而影響 DMD 顯示品質或是系統設定。相較之下,TCP具備可靠的錯誤更正與流量控制機制,能確保數據傳輸的準確性,但在 FPGA 端實作完整的 TCP 協議將增加硬體資源消耗與設計複雜度。 為了解決上述的問題,本研究提出一種適用於 FPGA 的類 TCP 協議,在 PC 端使用標準 TCP/IP 堆疊(無需任何特製驅動或應用程式修改)的同時,結合 10GbE(10 Gigabit Ethernet)與 SFP+(Small Form-factor Pluggable)高速光纖通訊介面,希望能在擁有高吞吐量的同時,提供基本的錯誤監測、封包回傳的機制,以確保整個系統的穩定性。 | zh_TW |
| dc.description.abstract | The Digital Micromirror Device (DMD) is widely used in applications such as projection displays and optical computing, demanding stringent requirements for real-time performance and high-throughput data transmission. In this thesis, data is transmitted from the computer via a socket to the FPGA (Field Programmable Gate Array), where it undergoes processing before being sent to the DMD through the FMC (Field-Programmable Mezzanine Card) interface for display. Selecting an appropriate communication protocol is crucial for efficient data transmission. UDP (User Datagram Protocol) and TCP (Transmission Control Protocol) under the Ethernet protocol are commonly used solutions, each with its own advantages and disadvantages in terms of transmission efficiency and error control capability.
In the Ethernet protocol, while UDP can provide high throughput, it lacks error control mechanisms, making it prone to data loss or errors, which can negatively impact the quality of the DMD display or system configuration. In contrast, TCP offers reliable error correction and flow control mechanisms, ensuring accurate data transmission. However, implementing a complete TCP protocol stack on an FPGA increases hardware resource consumption and design complexity. To address these challenges, this study proposes an FPGA-tailored TCP-like protocol that—while fully compatible with the PC’s standard TCP/IP stack (requiring no custom drivers or application changes)—leverages 10 GbE and SFP+ high-speed optical interfaces to achieve high throughput and incorporates essential error-detection and packet-retransmission mechanisms to ensure overall system stability. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2025-07-24T16:09:34Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2025-07-24T16:09:34Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | 致謝 i
摘要 ii 英文摘要 iv 目次 vii 圖次 xi 表次 xiii 縮寫列表 xiv 第一章 緒論 1 1.1 研究背景 1 1.2 研究動機 1 1.3 研究目的 3 1.4 本文架構 3 第二章 背景介紹及文獻回顧 4 2.1 DMD系統與DLP9000X[1]介紹 4 2.1.1 DMD工作原理 4 2.1.2 DMD微鏡陣列結構示意 5 2.1.3 DLP9000X主要規格 5 2.2 UDP與TCP傳輸協定 6 2.2.1 用戶資料報協定(UserDatagramProtocol, UDP) 6 2.2.1.1 高吞吐量與低延遲特性 6 2.2.1.2 可靠性與資料遺失的影響 7 2.2.2 傳輸控制協定(TransmissionControlProtocol, TCP) 7 2.2.2.1 TCP標頭格式(TCP Header Format) 8 2.2.2.2 TCP的連線建立與確認機制 9 2.2.2.3 順序號碼/確認號碼(ACK/SEQ Number)機制 11 2.2.2.4 在高吞吐量應用中的限制 12 2.2.3 UDP與TCP的比較與小結 13 2.3 FPGA與其在本研究中的角色 14 2.3.1 Advanced eXtensible Interface(AXI)[2] 15 2.3.2 FPGA在高速資料處理與通訊中的應用 17 2.3.3 FPGA在本研究中的角色與功能 18 2.4 10GbE通訊協定與傳輸技術[3] 20 2.4.1 10GbE(10-GigabitEthernet)概述 20 2.4.2 10GbE的技術架構 20 2.4.2.1 MAC層與verilog-ethernet[4] 21 2.4.2.2 PCS(Physical Coding Sublayer) 22 2.4.2.3 PMA(Physical Medium Attachment) 24 2.4.2.4 XGMII(10-Gigabit Media Independent Interface) 25 2.5 相關研究與類似專案回顧 27 2.5.1 基於PCIe的FPGA網路加速專案 28 2.5.2 基於Ethernet的FPGA網路傳輸專案 28 2.5.3 本研究與現有專案的異同 28 2.5.4 小結 29 第三章 FPGA系統設計 30 3.1 硬體平台 30 3.2 系統架構概述 31 3.2.1 系統設計目標與需求 31 3.2.2 整體系統架構圖 32 3.3 模組設計與實現 34 3.3.1 類TCP協議層設計 34 3.3.1.1 封包格式與協議修改 34 3.3.1.2 指令封包與資料封包類型 35 3.3.2 TCP_RX與TCP_TX狀態機設計 35 3.3.2.1 TCP_RX狀態機 35 3.3.2.2 TCP_TX狀態機 36 3.3.3 核對和與錯誤重傳機制 37 3.3.4 實作後資源消耗比較(TCP_inst vs. UDP_inst) 38 3.4 小結 38 第四章 實驗結果與討論 39 4.1 實驗平台與測試環境 39 4.2 實驗方法 40 4.2.1 Socket傳輸速率測試 41 4.2.2 可調整丟包率測試 42 4.3 實驗結果 43 4.3.1 Socket傳輸速率測試 43 4.3.2 可調整丟包率測試 45 4.4 結果討論 46 4.5 小結 47 第五章 結論與未來展望 48 5.1 結論 48 5.2 未來展望 49 參考文獻 51 | - |
| dc.language.iso | zh_TW | - |
| dc.subject | 使用者資料報協定 | zh_TW |
| dc.subject | 現場可程式化邏輯陣列 | zh_TW |
| dc.subject | 數位微鏡裝置 | zh_TW |
| dc.subject | 傳輸控制協定 | zh_TW |
| dc.subject | TCP | en |
| dc.subject | DMD | en |
| dc.subject | FPGA | en |
| dc.subject | 10GbE | en |
| dc.subject | UDP | en |
| dc.title | 基於FPGA的10GbE高吞吐量類TCP協議—應用於即時大量影像資料傳輸 | zh_TW |
| dc.title | A High-Throughput 10 GbE TCP-like Protocol on FPGA for Real-Time, Massive Data Transmission | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 113-2 | - |
| dc.description.degree | 碩士 | - |
| dc.contributor.oralexamcommittee | 陳彥廷;謝宏昀;彭盛裕 | zh_TW |
| dc.contributor.oralexamcommittee | Yen-Ting Chen;Hung-Yun Hsieh;Sheng-Yu Peng | en |
| dc.subject.keyword | 數位微鏡裝置,現場可程式化邏輯陣列,使用者資料報協定,傳輸控制協定, | zh_TW |
| dc.subject.keyword | DMD,FPGA,10GbE,UDP,TCP, | en |
| dc.relation.page | 52 | - |
| dc.identifier.doi | 10.6342/NTU202501253 | - |
| dc.rights.note | 未授權 | - |
| dc.date.accepted | 2025-07-14 | - |
| dc.contributor.author-college | 工學院 | - |
| dc.contributor.author-dept | 工程科學及海洋工程學系 | - |
| dc.date.embargo-lift | N/A | - |
| 顯示於系所單位: | 工程科學及海洋工程學系 | |
文件中的檔案:
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-113-2.pdf 未授權公開取用 | 2.72 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。
