隨著網路科技蓬勃發展，使用者的頻寬使用量也越來越大。對於即時影音串流系統來說，雖然有助於它的發展以提供更高解析度的影像畫質，但相對地，它們付出相當大的花費在頻寬成本上。而藉由P2P架構，使用者可以分享他們頻寬以減少系統本身的負擔。但由於須在時限內將影音串流傳送到使用者，P2P拓樸便成為系統中相當重要的一環，一個強壯及良好的拓樸能給予使用者好的服務品質，而一個脆弱拓樸會導致頻繁的拓樸變動，進而影響使用者的服務品質，降低使用者體驗。
在本篇論文中，我們研究著重於如何建立一個良好的拓樸並維護它。在三個層面限制下，提供良好的影像品質。第一個層面是盡可能節省頻寬成本，充分利用使用者的頻寬資源；第二個層面是在基於一個保證影像延遲底下，再盡可能降低使用者的影像延遲；最後一個層面是盡可能保持住良好的拓樸，避免良好的拓樸崩壞。基於這三個層面，我們實作出一套即時影音串流系統，並設計一系列的演算法強化此系統。
實驗方面，我們在PlanetLab上做一系列的實驗以驗證所提出的演算法。除此之外，我們更在真實網路環境測試我們的系統，使用者遍及全國各地。實驗結果顯示我們的演算法能強化系統，更證明我們的系統在真實環境之可行性。

With the development of Internet, the requirement of bandwidth usages increases. Live video streaming systems are favorable to provide higher resolution video; however, they undertake more bandwidth costs. By means of P2P architecture, users share their bandwidth to save bandwidth resource on servers. Since the data information has to be disseminated to all of users with time limit, the topology of P2P systems plays a considerable role. A well-structured topology can provide users better quality of services while a fragile topology decreases quality of services even user experience.
In this thesis, our research aims at how to construct a robust topology and how to maintain it with the following limitations. First, the system should save the bandwidth resource as more as possible and exploit the upload bandwidth from users. Second, despite the guaranteed latency, the system should minimize the delay of video. Third, the system prevents the topology from fragments. Based on these limitations, we implement a P2P based live streaming system, and several proposed algorithms are designed to improve the system.
The proposed algorithms are verified in the experiments performed on PlanetLab. Furthermore, we perform nationwide experiments in Taiwan. The experimental results show the superior performance of the algorithms and the feasibility.

Contents

Chapter 1 Introduction 1
Chapter 2 Related Works 3
2.1 Tree-Push Topology 3
2.2 Mesh-Pull Topology 4
2.3 Mesh-Push Topology 5
2.4 Hybrid Tree-Mesh Topology 6
Chapter 3 System Architecture 8
3.1 System Overview 8
3.1.1 Sources 8
3.1.2 Media Server 9
3.1.3 Manager Server 9
3.1.4 PK Server 9
3.1.5 Log Server 10
3.1.6 Channel Network 10
3.2 Chunk Format 10
3.3 Streaming 11
3.4 Time Synchronization 12
3.5 Source Delay 13
3.6 Rescue 14
3.7 Two-way Connect 15
Chapter 4 Algorithms 17
4.1 Queue Scheduling 18
4.2 ESD Parent Selection 19
4.3 Peer Loading Detection 23
4.4 Rescue Grouping 26
Chapter 5 Experimental Results 30
5.1 Experiments on PlanetLab 30
5.1.1 Experiments on Different Parent Selections 31
5.1.2 Experiments on Peer Loading Detection 35
5.1.3 Experiments on Rescue Grouping 40
5.2 Nationwide Experiments 42
Chapter 6 Conclusion and Future Works 46
Figure List

Fig. 2 1 Tree-Push Topology 4
Fig. 2 2 Mesh-Pull Topology 5
Fig. 2 3 Mesh-Push Topology 6
Fig. 2 4 Hybrid Tree-Mesh Topology 7
Fig. 3 1 System Architecture 8
Fig. 3-2 Chunk Format 10
Fig. 3-3 Streaming Flow 12
Fig. 3 4 Time Synchronization 12
Fig. 3 5 Source Delay of Chunk k 14
Fig. 3-6 Incoming Flow Blocked by Firewall 15
Fig. 3-7 Procedure of Two-Way Connect 16
Fig. 4 1 The Block Diagram of the Algorithms in Peer Life Cycle 18
Fig. 4 2 The Queueing Model 19
Fig. 4 3 The Bandwidth-N(q) Model 20
Fig. 4 4 The Illustration of Estimated Source Delay 22
Fig. 4 5 The Illustration of Single Node Failure Problem 27
Fig. 5-1 Nodes Distribution on PlanetLab 30
Fig. 5-2 Peer Source Delay Distribution with Different Selection Policies 33
Fig. 5-3 Performance with Different Selection Policies 34
Fig. 5-4 Comparison of Peer Loading Detection 36
Fig. 5-5 Standard Deviation of Bandwidth Influenced by α 37
Fig. 5-6 Server Peak Bandwidth Influenced by α 37
Fig. 5-7 Number of Rescues Influenced by α 38
Fig. 5-8 Mean Source Delay Influenced by α 39
Fig. 5-9 Continuity Influenced by α 39
Fig. 5-10 Server Bandwidth Saving Ratio Influenced by α 39
Fig. 5-11 Comparison of Rescue Grouping 41
Fig. 5-12 Performance of Rescue Grouping 42
Fig. 5-13 Platform for Live Streaming 43
Fig. 5-14 Comparison of Two-way Connect and STUN 44
 
Table List

Table 4 1 Peer Substream-State 21
Table 4 2 Definitions and Symbols 24
Table 4 3 Pseudo Code of Peer Loading Detection 25
Table 4 4 Pseudo Code of Source Delay Detection 28
Table 5 1 Nationwide Experiments using STUN 44
Table 5 2 Nationwide Experiments using Two-way Connect 45

[1] PlanetLab. [Online]. Available: https://www.planet-lab.org/
[2] YouTube. [Online]. Available: https://www.youtube.com/
[3] PPStream. [Online]. Available: http://www.pps.tv/
[4] CoolStreaming. [Online]. Available: http://webtv.coolstreaming.us/
[5] H. Hongmu, H. Jie, and Z. Cuihua, "A hybrid P2P overlay network for high efficient search," in Information and Financial Engineering (ICIFE), 2010 2nd IEEE International Conference on, Chongqing, China, Sept. 2010, pp. 241-245.
[6] R. Lo Cigno, A. Russo, and D. Carra, "On some fundamental properties of P2P push/pull protocols," in Communications and Electronics, 2008. ICCE 2008. Second International Conference on, June 2008, pp. 67-73.
[7] K. Chee Yik, H. Poo Kuan, and T. Choo-Yee, "Efficient Hybrid Push-Pull Based P2P Media Streaming System," in Parallel and Distributed Systems (ICPADS), 2011 IEEE 17th International Conference on, Tainan, Taiwan, Dec. 2011, pp. 735-740.
[8] S. Banerjee, B. Bhattacharjee, and C. Kommareddy, "Scalable application layer multicast," SIGCOMM Comput. Commun. Rev., vol. 32, no. 4, pp. 205-217, Oct. 2002.
[9] Y. H. Chu, S. G. Rao, S. Seshan, and Z. Hui, "A case for end system multicast," Selected Areas in Communications, IEEE Journal on, vol. 20, no. 8, pp. 1456-1471, Oct. 2002.
[10] R. Dongni, W. Wong, and S. H. G. Chan, "Toward continuous push-based P2P live streaming," in Global Communications Conference (GLOBECOM), 2012 IEEE, Anaheim, CA, Dec. 2012, pp. 1969-1974.
[11] Y. Tang Yun, L. Jian-Guang, M. Zhang Meng, Z. Meng, and Y. Shi-Qiang, "Deploying P2P networks for large-scale live video-streaming service [Peer-to-Peer Multimedia Streaming]," Communications Magazine, IEEE, vol. 45, no. 6, pp. 100-106, June 2007.
[12] H. Xiaojun, L. Yong, and K. W. Ross, "IPTV over P2P streaming networks: the mesh-pull approach," Communications Magazine, IEEE, vol. 46, no. 2, pp. 86-92, Feb. 2008.
[13] S. Yi-Yu, L. Pin-Chuan, and K. Ching-Chun, "Efficient startup segment selection and scheduling for mesh-pull peer-to-peer live streaming system," in Consumer Electronics (GCCE), 2012 IEEE 1st Global Conference on, Tokyo, Japan, Oct. 2012, pp. 450-451.
[14] W. Mea and L. Baochun, "Lava: A Reality Check of Network Coding in Peer-to-Peer Live Streaming," in INFOCOM 2007. 26th IEEE International Conference on Computer Communications. IEEE, Anchorage, AK, May 2007, pp. 1082-1090.
[15] W. Mea and L. Baochun, "R2: Random Push with Random Network Coding in Live Peer-to-Peer Streaming," Selected Areas in Communications, IEEE Journal on, vol. 25, no. 9, pp. 1655-1666, Dec. 2007.
[16] C. Laizhong, J. Yong, and W. Jianping, "Optimizing push scheduling algorithm based on network coding for mesh Peer-to-Peer live streaming," in Communications (ICC), 2012 IEEE International Conference on, Ottawa, ON, June 2012, pp. 2075-2080.
[17] A. Fiandrotti, A. M. Sheikh, and E. Magli, "Towards a P2P videoconferencing system based on low-delay network coding," in Signal Processing Conference (EUSIPCO), 2012 Proceedings of the 20th European, Bucharest, Aug. 2012, pp. 1529-1533.
[18] W. Feng, X. Yongqiang, and L. Jiangchuan, "mTreebone: A Hybrid Tree/Mesh Overlay for Application-Layer Live Video Multicast," in Distributed Computing Systems, 2007. ICDCS '07. 27th International Conference on, Toronto, ON, June 2007, pp. 49-49.
[19] R. Dongni, Y. T. H. Li, and S. H. G. Chan, "Fast-Mesh: A Low-Delay High-Bandwidth Mesh for Peer-to-Peer Live Streaming," Multimedia, IEEE Transactions on, vol. 11, no. 8, pp. 1446-1456, Dec. 2009.
[20] Open Broadcaster Software. [Online]. Available: https://obsproject.com
[21] Real-Time Messaging Protocol. [Online]. Available: http://www.adobe.com/devnet/rtmp.html/.
[22] Red5 Media Server. [Online]. Available: https://github.com/Red5
[23] Flash Video (FLV). [Online]. Available: http://zh.wikipedia.org/wiki/Flash_Video
[24] D.L. Mills, “Network Time Protocol,” RFC 958, Internet Engineering Task Force, September 1985. [Online]. Available: https://www.ietf.org/rfc/rfc958.txt
[25] J. Rosenberg, J. Weinberger, C. Huitema, and R. Mahy “STUN - Simple Traversal of User Datagram Protocol (UDP) Through Network Address Translators (NATs),” RFC 3489, Internet Engineering Task Force, March 2003. [Online]. Available: https://www.ietf.org/rfc/rfc3489.txt
[26] UDP-based Data Transfer (UDT). [Online]. Available: http://udt.sourceforge.net/
[27] X. Xin, S. Yuanchun, Z. Baopeng, and G. Yuan, "OCals: A Novel Overlay Construction Approach for Layered Streaming," in Communications, 2008. ICC '08. IEEE International Conference on, Beijing, China, May 2008, pp. 1807-1812.
[28] A. P. C. da Silva, E. Leonardi, M. Mellia, and M. Meo, "A Bandwidth-Aware Scheduling Strategy for P2P-TV Systems," in Peer-to-Peer Computing, 2008. P2P '08. Eighth International Conference on, Aachen, Sept. 2008, pp. 279-288.
[29] J. Ghosha, W. Miao, X. Lisong, and B. Ramamurthy, "Variable neighbor selection in live peer-to-peer multimedia streaming networks," in Broadband Communications, Networks and Systems, 2008. BROADNETS 2008. 5th International Conference on, London, Sept. 2008, pp. 344-346.
[30] W. Lei, Z. Dengyi, and Y. Hongyun, "Qos-Awareness variable neighbor selection for mesh-based P2P live streaming system," in Information Science and Technology (ICIST), 2013 International Conference on, Yangzhou, China, March 2013, pp. 1197-1201.
[31] C. Laizhong, J. Yong, and W. Jianping, "Employing QoS Driven Neighbor Selection for Heterogeneous Peer-to-Peer Streaming," in Communications (ICC), 2011 IEEE International Conference on, Kyoto, Japan, June 2011, pp. 1-6.
[32] H. Tai-Hua, H. Ming-Hung, and M. Yu-Ben, "Adaptive and Efficient Peer Selection in Peer-to-Peer Streaming Networks," in Parallel and Distributed Systems (ICPADS), 2011 IEEE 17th International Conference on, Dec. 2011, pp. 753-758.
[33] L. Lima, M. Dalai, R. Leonardi, P. Migliorati, R. Bernardini, and R. Rinaldo, "Optimal Rate Allocation for P2P Video Streaming," Selected Areas in Communications, IEEE Journal on, vol. 31, no. 9, pp. 200-213, Sept. 2013.
[34] L. Xiaoqun, C. Laizhong, and X. Shutao, "An effective transmission scheduling mechanism with network coding for adaptive P2P streaming," in Multimedia Technology (ICMT), 2011 International Conference on, Hangzhou, China, July 2011, pp. 4820-4824.
[35] A. Carta, M. Mellia, M. Meo, and S. Traverso, "Efficient Uplink Bandwidth Utilization in P2P-TV Streaming Systems," in Global Telecommunications Conference (GLOBECOM 2010), 2010 IEEE, Miami, FL, Dec. 2010, pp. 1-6.
[36] R. Birke, C. Kiraly, E. Leonardi, M. Mellia, M. Meo, and S. Traverso, "A delay-based aggregate rate control for P2P streaming systems," Computer Communications, vol. 35, no. 18, pp. 2237-2244, Nov. 2012.