隨著多媒體技術與網路技術的蓬勃發展，即時性的網路視頻服務已經成為目前主流的數位內容服務項目之一，因此如何提升多媒體網路應用服務於網路架構穩定性乃為當今重要議題之一，由於多媒體網路服務運作時特別注重「即時性」與「穩定性」，尤其是在多位使用者使用的情形下，因此為了達到「網路拓撲最佳化」的目標，必須以「可用頻寬估測值」來定義網路頻寬能力不同的終端用戶，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，讓使用者能夠收看本身網路能力所能夠收看的多媒體節目之最高品質，並讓視頻在收播的過程可以能順暢與穩定，提升整體用戶收視之品質。本論文提出了一種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)，並且加入判斷探測封包的單向路徑延遲(One-Way Delay)轉折趨勢與封包遺失率(Packet-Loss Rate)的概念，使其成為計算可用頻寬估測演算法的依據，最後做「階層式」地收斂頻寬估測範圍，求出最後的可用頻寬估測值，藉此可提升在多媒體網路即時環境下的估測速度及其精確度。本論文利用整合了跨平台封包發送工具-iperf的頻寬估測系統，以Visual C# 2005程式語言分別實作三種頻寬估測技術-Pathload、PathQuick與本實驗室開發的HABE探測技術，將上述三種頻寬估測方法的探測封包延遲統計、探測速度、探測封包總量與探測準確度以數據量表顯示來做比較，藉此證明HABE頻寬探測機制的技術改良性，最後利用HABE探測出來的可用頻寬值提供給IPTV系統作為辨別其連網路線種類(有線、WiFi連線、3G連線)的依據。

With the advance of multimedia communication technologies, real-time video streaming services becomes one of the major digital content services. How to improve the stability for real-time video streaming when many users are requesting the service is very important. To achieve this goal, controlling the network connection topology in an optimal way may help, under which available bandwidth estimation procedure is required. User peers can then be classified by their processing speed and bandwidth capability. The peer interconnection tree can be constructed in an optimal way according to peer capability such that high capability peers are interconnected in upper level and vise verse. Under this interconnection strategy, the overall end user can be served with best video transmission and reconstruction quality. We proposed to estimate the available bandwidth between two connected peers in a hierarchical way, abbreviated as Hierarchical Available Bandwidth Estimation, HABA, to speed up the estimation procedure. The major probing mechanism comprises: the speed rate model that provides several probing packet train transmission rate and the procedure that can judge the one way delay increase/decrease and packet loss rate, such that they can be used to estimate bandwidth. This probing mechanism is carried out in a hierarchical way to speed up the estimation procedure for accurate bandwidth estimation. The estimated bandwidth, together with peer CPU power, can then be utilized to help construct an optimal broadcasting tree in an optimal approach. The bandwidth estimation methods were developed based on the Visual C# 2005 platform and three different methods, Pathload, PathQuick and the proposed HABA are implemented for comparisons. The well-known bandwidth estimation tool, iperf, is used as the benchmark for performance evaluation. Experiments showed that the proposed HABA method outperform the others in bandwidth estimation accuracy. This proposed HABA method can be utilized for the IPTV system to provide suitable QoS for different users connected through wired, WiFi or 3G networks.

[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] 陳克任, “多媒體通訊”, 1999年出版
[4] http://letsblogblogger.blogspot.tw/2011/06/short-introduction-to-user-datagram.html
[5] Douglas E. Comer 原著, 張智聖與陳伯偉編譯, “TCP/IP互連網路(第五版)”
[6] 黃嘉輝, “Visual C# 2005網路程式設計”, 2007年出版
[7] 康廷數位, “.Net網路與I/O技術手冊(第二版)”, 2010年出版
[8] 林豐仁, “具動態頻寬估測之網路視訊電話實作”, 國立台北科技大學電機工程系碩士學位論文, 2008
[9] 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
[10] 韋安明，王洪波，林宇，程時端, “IP網頻寬測量技術研究與進展”電子學報, VOl.34, No.7, July 2006
[11] 林華毅, “具使用者頻寬估測之IPTV系統設計與實作”, 2011
[12] Ravi Prasad and Constantinous Dovroils, “Bandwidth estimation - metrics, measurement techniques, and tools,” IEEE Network November/December Journal , 2003
[13] N. Hu, and P. Steenkiste, “Evaluation and characterization of available bandwidth probing techniques,” IEEE JSAC, vol. 21, no. 6, pp. 879-894, August 2003.
[14] 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.
[15] V. Riberiro et al., “Multifractal cross-traffic estimation,” ITC Specialist Seminar on IP Traffic Measurement, Modeling, and Management, September 2000.
[16] M. Zhang, C. Luo and J. Li, “Estimation available bandwidth using multiple overloading streams,” IEEE ICC2006, pp. 495-502, June 2006.
[17] M. Jain and C. Dovrolis, “End-to-End available bandwidth: measurement methodology, dynamics, and relation with TCP Throughput,” IEEE/ACM Trans. on Networking, Vol. 11, No. 4, pp. 537-549, August 2003.
[18] http://pirlwww.lpl.arizona.edu/resources/guide/software/iperf/
[19] http://www.wcisd.hpc.mil/nuttcp/Nuttcp-HOWTO.html
[20] 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.
[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 Laboratories, NEC Corporation, Japan, 2010.