臺灣博碩士論文加值系統

近幾年多媒體信號處理技術進步與行動網路（Cellular Networks）蓬勃發展，透過行動網路即時串流影音媒體觀看已經成為重要的網路服務之一。在頻寬不穩定及異質網路環境下如何提供穩定收視服務為最關鍵的技術項目之一。因為即時影音串流服務注重「即時性」、「穩定性」及「高解析度」，部份多媒體系統會以「可用頻寬估測」(Available Bandwidth Estimation, ABE)來測量終端用戶的網路頻寬，並依照其可用頻寬估測值傳送適當位元編碼率的多媒體影音內容，如此用戶根據可用頻寬接收到最合適之多媒體影音服務。相較於有線網路，行動網路容易受環境影響而導致可用頻寬變動，經由行動網路接收即時影音串流服務時，若頻寬瞬間由高變低且系統持續傳送高位元編碼率之影音內容，則會使即時影音串流無法穩定播放。若系統可以在短時間內探測可用頻寬變化，並據以轉碼以提供相應位元率之影音碼流，則可避免影片播放中斷，但要如何在短時間內探測出可用頻寬為一具挑戰性的關鍵技術。目前大部份的可用頻寬估測工具主要專注於精準度，且需要較長時間進行頻寬估測，且會發送大量封包會影響其餘網路服務。本論文提出適用於多媒體網路即時服務的「階層式可用頻寬估測」（Hierarchical Available Bandwidth Estimation, HABE）技術來探測終端用戶之頻寬能力，其探測機制主要是採用包含多種探測速率之探測封包序列（Probing Packets Train）的速率模型（PRM），觀察單向路徑延遲（One-Way Delay）轉折趨勢做為可用頻寬估測演算法的依據，並可經由多次收斂來求得更精確的可用頻寬。HABE盡可能減少收斂次數以降低探測頻寬所需時間，以提升頻寬估測的速度，使其適用於在即時性多媒體系統。

With the advance of multimedia processing technologies and wireless networks, watching live video streaming through the wireless network has become one of the most popular network application services. However, due to instable wireless network transmission, how to provide stable media streaming services is critical for this application. As the live video streaming service requires real-time and stability transsmisoin to provide high-quality perception, some advanced multimedia streaming service systems execute an “available bandwidth estimation (ABE)” procedure to estimate available bandwidth of end users, such that the system can transcode and provide the most matched bitrate video according to estimatied ABE. Users can perceive stable and most suitable video quality with the help of this quick available bandwidth estimation. As compared to the wired networks, the wireless networks bandwidth would be sensitive to environmental factors, under which the live video streaming would not be smooth if the end-to-end available bandwidth was not stable. Video playback interruption due to unstable wireless network transmission can be avoided if the media streaming system can quickly detect changes in client bandwidth, and transmit bitrate matched video content. How to pre-estimate the ABE in a short period of time is critical. Current ABE tools emphasizes precision and require a much longer period of time. They would also send a large number of probing network packets and suffer the heavy network loading problem. We proposed a Hierarchica Available Bandwidth Estimation (HABE) method that can estimate the ABE in a shorter period of time. The HABE adopted the probing rate model (PRM) that utilizes multi packet rates as a probing packet train to observ their up-turn of one-way dely (OWD) times to estimate the ABE. By performing this procedure hierarchically, the ABE can be estimated in a much shorter period of time. Experiments justified the proposed HABE capability in providing fast and accurate ABE results.

摘要 I
ABSTRACT II
誌謝 III
目錄 IV
圖目錄 VI
第一章 緒論 1
1.1 前言 1
1.2 研究背景與動機 2
1.3 研究項目與方法概述 4
1.4 論文架構 5
第二章 背景知識與相關研究探討 6
2.1 多媒體傳輸網路簡介 6
2.1.1 多媒體網路傳輸協定簡介 6
2.1.1.1 可靠性資料流傳輸控制協議（Transmission Control Protocol ,TCP） 7
2.1.1.2 使用者數據包協議（User Datagram Protocol ,UDP） 8
2.2 可用頻寬之背景知識簡介 9
2.2.1 可用頻寬定義簡介 11
2.2.2 單向路徑延遲介紹（One-Way-Delay） 13
2.2.3 動態單向路徑延遲議題（Dynamic One-Way-Delay） 14
2.2.4 可用頻寬技術分類 17
2.2.4.1 間隙模型（Probe Gap Model, PGM ） 18
2.2.4.2 速率模型（Probe Rate Model, PRM ） 19
2.3 現有可用頻寬估測工具之介紹 21
2.3.1 Iperf 21
2.3.2 Nuttcp 23
2.3.3 Dr. Speed 24
第三章 階層式可用頻寬估測 27
3.1 現有可用頻寬估測技術之分析 28
3.1.1 PathLoad 28
3.1.2 PathQuick 31
3.2 階層式可用頻寬估測系統架構簡介 33
3.3 階層式可用頻寬估測技術之概念簡介 34
3.3.1 符合多即時多媒體應用之頻寬估測 34
3.3.2 探測封包序列設計 36
3.4 階層式可用頻寬估測演算法之探測流程 41
3.4.1 HABE頻寬探測流程圖說明 41
3.4.1.1 HABE流程說明 - ① Server與Client連線建立與時間同步 42
3.4.1.2 HABE流程說明 - ② Server流程詳述 44
3.4.1.3 HABE流程說明 - ③ Client流程詳述 46
3.4.2 藉由單向路徑延遲趨勢之轉折點判斷可用頻寬 48
3.5 階層式可用頻寬估測技術之理論總結 51
第四章 系統實作與數據結果 52
4.1 NS2頻寬估測精確度模擬 52
4.2 階層式可用頻寬估測系統實作與功能說明 54
4.2.1 Server Mode 55
4.2.2 Client Mode 56
4.2.3 階層式可用頻寬估測系統運作過程說明 57
4.3 實際量測數據比較 59
4.3.1 無線AP頻寬限流 59
4.3.1.1 無線AP頻寬限流 – 估測頻寬準確度分析 59
4.3.1.2 無線AP頻寬限流 – 頻寬估測時間與探測封包總額分析 61
4.3.2 跨網域有線網路 62
4.3.2.1 跨網域有線網路 - 估測頻寬準確度分析 63
4.3.2.2 跨網域有線網路 - 頻寬估測時間與探測封包總額分析 64
4.3.3 行動網路 65
4.3.3.1 行動網路 - 估測頻寬準確度分析 65
4.3.3.2 行動網路 - 頻寬估測時間與探測封包總額分析 67
4.4 實際量測數據總結 69
第五章 結論與未來研究探討 70
5.1 結論 70
5.2 未來研究探討 71
參考文獻 72

[1]A. Yarali and A. Cherry, “Internet Protocol Television（IPTV）,” IEEE TENCON, pp. 1-6,2005.
[2]Chao Liang and Young Liu, “Topology optimization in multi-tree based p2p streaming system”, IEEE International Conference on Tools with Artificial Intillengence, 2009.
[3]Behrous A. Foruzan, Date Communications and Networking 5e , McGraw Hill Education
[4]Douglas E. Comer 原著, 張智聖與陳伯偉編譯, “TCP/IP互連網路（第五版）”
[5]James F. Kurose, Keith W. Ross原著, 吳家榮與黃彩嵐編譯, “電腦網際網路（第三版）”, 2006
[6]韦安明，王洪波，林宇，程时端, “IP网带宽测量技术研究与进展”电子学报, VOl.34, No.7, July 2006
[7]林華毅, “具使用者頻寬估測之IPTV系統設計與實作”, 2011
[8]Ravi Prasad and Constantinous Dovroils, “Bandwidth estimation - metrics, measurement techniques, and tools,” IEEE Network November/December Journal , 2003
[9]CARTER R.L., CROVELLA M.E. , “Measuring bottleneck link speed in packet- switched networks,” Technical Report, BU-CS-96-006, Boston University, 1996
[10]DOVROLIS C., RAMANATHAN P., MOORE D., “What do packet dispersion techniques measure?,” in Proc. IEEE INFOCOM, 22–26 April 2001, vol. 2, pp. 905–914
[11]LAI K., BAKER M., “Measuring link bandwidths using a deterministic model of packet delay,” Proc. ACM SIGCOMM 2000, August 2000, pp. 283–294
[12]M. Jain and C. Dovrolis, “Pathload: a measurement tool for end-to-end available bandwidth,” Passive and Active Measurements Workshop, Fort Collins, CO, March 2002.
[13]Chobanyan A., Mutka M., Cen Z., Xi N. , “One way delay trend detection for available bandwidth measurement,” Proc. IEEE GLOBECOM Conf., 2005
[14]Melander B., Bjorrkman M., Gunningberg P., “First-comefirst-served packet dispersion and implications for TCP,” Proc. IEEE GLOBECOM-02, vol. 3, pp. 2170–2174, Nov. 2002.
[15]Strauss J., Katabi D., Kaashoek F., “A measurement study of available bandwidth estimation tools,” The Internet Measurements Conf., Florida, 2003, pp. 39–44
[16]N. Hu, and P. Steenkiste, “Evaluation and characterization of available bandwidth probing techniques,” IEEE JSAC, vol. 21, no. 6, pp. 879-894, August 2003.
[17]J. Strauss, D. Katabi, and F. Kaashoek, “A measurement study of available bandwidth estimation tools,” ACM SIGCOMM Conference on Internet Measurements, pp. 39-44, Florida, Oct. 2003.
[18]V. Riberiro et al., “Multifractal cross-traffic estimation,” ITC Specialist Seminar on IP Traffic Measurement, Modeling, and Management, September 2000.
[19]M. Zhang, C. Luo and J. Li, “Estimation Available Bandwidth Using Multiple Overloading Streams,” IEEE ICC2006, pp. 495-502, June 2006.
[20]M. Jain and C. Dovrolis, “End-to-End available bandwidth: measurement methodology, dynamics, and relation with TCP throughput,” IEEE/ACM Trans. Networking, vol. 11, no. 4, pp. 537-549, Aug. 2003.
[21]V. J. Ribeiro et al., “PathChirp: efficient available bandwidth estimation for network paths,” PAM Workshop, 2003.
[22]T. Oshiba and K. Nakajima, “Quick end-to-end available bandwidth estimation for QoS of real-time multimedia communication,” Service Platforms Research Labs., NEC Corp., Japan, 2010.
[23]iPerf, Retrieved August 23, 2014, from https://iperf.fr/
[24]Nuttcp, Retrieved August 23, 2014, from http://fasterdata.es.net/performance-testing/network-troubleshooting-tools/nuttcp/
[25]NiNet連線速率測試, Chunghwa Telecom Co. Retrieved August 24, 2014, from http://speed.hinet.net/
[26]Wireshark, Retrieved August 23, 2014, from https://www.wireshark.org/
[27]J. Rosenberg, R. Mahy, P.Matthews, D.Wing, “Internet protocol, version 6 （IPv6） specification,” December 1998. RFC 2460
[28]Dynamic adaptive streaming over HTTP（DASH）. Retrieved August 23, 2014, from http://dashif.org/
[29]Hu N., Steenkiste P, “Evaluation and characterization of available bandwidth probing techniques,” IEEE J. Sel. Areas Commun., vol. 21, no. 6, pp. 879 –894, 2003.
[30]Shriram A., Murray M., Hyun Y., et al, “Comparison of public end-to-end bandwidth estimation tools on highspeed links,” Proc. Passive and Active Measurement Workshop, pp. 222–235, 2005.