隨著網際網路的進步與成長，有越來越多的應用將會在網路的協助下實現。在影音應用的發展上，除了過往影音資源分享的需求，現今的使用者也期待可以獲得即時且高品質的影音傳輸服務。目前市面上已經有一些服務供應商提供了此類的服務，例如Youtube,Twitch。他們往往利用基於client/server的架構來提供給所有希望獲取服務的使用者影音資源。此舉雖然可以讓使用者們有機會可以在較低的延遲下獲取影音串流，但隨之而來對於網路頻寬的需求卻也會使的服務供應商需要花費大量的資本在網路資源的花費上。因此崛起於2000年左右的P2P(Peer to Peer) 系統架構開始展現其優勢，利用其本身對於伺服器較低量的需求，以及易於部署的特性，給予了即時影音串流服務一個新的選項。
但由於P2P架構先天上的設計，在P2P網路中大部分使用者的資源都必須由在同個P2P網路中的其他使用者提供，這將會使的使用者得到的影音品質無法維持恆定，而在當中也啟發了一個新的問題。當使用者分佈的區域極廣，該如何調整彼此間的連線，以改善即時影音傳輸的服務品質?對此我們提出了一些想法，利用每個使用者到其他使用者的距離，我們有機會因此而將所有的使用者做區域化的分類，由這個分類我們可以得出一個對於系統中所有使用者有較佳品質表現的連線關係。搭配上一些在拓墣上的調整方式，我們可以減少資料在送達接收者時的傳輸路徑，並讓所有的使用者都可以有較佳以及較穩定的資料傳輸品質。
奠基於已有的即時影音串流系統，LLP2P，我們將這些想法實作在其之上，並藉由在planetLab network上面的量測，我們證明了使用了上述想法，可以讓位於世界各地的使用者都可以因為區域化的關係而盡量的由位於自己附近的使用者的服務。並因而獲取到較佳的影音觀看品質。

With the advance and development in the networks, more and more applications are realized with the help of Internet networking. In the progress of the video applications, users nowadays are looking forward to being served with a high-quality real time video streaming besides the existing video file sharing services. There are already some services concerning video streaming being put on the shelf by the service providers, for instance, YouTube and Twitch. They adopt the Client/Server based architecture to serve the users who want to enjoy the real time video streaming. Though the architecture allows the users to obtain the video streaming with low latency, it comes with the demands for the network bandwidth, which is costly for the service providers. Therefore, the P2P (Peer-to-Peer) architecture, which is introduced to the world near the 2000s, becomes quite popular for its few requirements for the central servers support and its easy deployment. Accordingly, P2P provides a new option for the video streaming applications.
However, the nature of the P2P architecture requires most of the users in the system to get involved in the video streaming transmission, which results in an unstable video streaming quality for the users. Besides, there is also another problem arising. Since the users come from all around the world, how to adjust the topology so as to improve the video streaming quality? In this thesis, we suggest some measures to classify the users into groups based on their locations. With the help of the localization mechanism, we are able to obtain a fine topology for the system. Combined with several topology adjustment methods for the delivery trees of the P2P network, the data transmission path for the receivers will be reduced, which enables them to enjoy a fine and stable video quality experience.
On the basis of an existent live video streaming system, LLP2P, the proposed thoughts will be implemented on it. By the experiments on the planetLab network, we have proved that with the help of those methods which we suggest, users across the globe can experience a better video quality by being served with the users nearby, thanks to the advantage of localization.

Chapter 1 Introduction 1
Chapter 2 Related Works 4
2.1. Live video streaming system 4
2.2. Peer to peer topology 4
2.3. K-nearest Neighbors Algorithm (KNN) 7
2.4. Traffic localization 7
2.5. Topology construction 7
2.6. Network coding 8
2.7 Commercial programs 8
Chapter 3 System architecture 10
3.1 Components overview 11
3.1.1 Media server 11
3.1.2 Manager server 11
3.1.3 The PK-server and the P2P media server 11
3.1.4 Client 12
3.1.5 Log server 12
3.1.6 Channel 12
3.1.7 Sub-stream 12
3.2 Content providing 13
3.3 Peer joining 13
3.4 Peer maintenance 14
Chapter 4 System Implementation 15
4.1 The RTT grouping mechanism 15
4.1.1 Background 15
4.1.2 RTT value retrieval 15
4.1.3 RTT group retrieval 16
4.2 Topology adjustment algorithms 17
4.2.1 Background 17
4.2.3 Downward relocation algorithm 18
4.2.2 Upward relocation algorithm 20
4.2.4 The combination 22
4.3 Merge algorithm 23
4.3.1 Background 23
4.3.2 Selection of parents 24
4.3.3 Proper seed number algorithm 25
4.3.4 Operation time 26
4.3.5 The combination 27
4.4 Delivery path adjustment 27
4.4.1 Background 27
4.4.2 Delivery path adjustment 27
4.5 Early rescue mechanism 28
Chapter 5 Performance Evaluation 30
5.1 Experiment spec 30
5.2 RTT grouping mechanism 30
5.2 Delivery path adjustment 32
5.3 Local topology adjustment 35
5.4 Tests with different guaranteed delay 38
5.5 Tests with different RTT_GROUP_BOUND 44
Chapter 6 Conclusion and Future works 51
References 53

[1] YouTube. [Online] Available: https://www.youtube.com/
[2] PPStream. [Online] Available: http://www.pps.tv/
[3] PPTV. [Online] Available: http://www.pptv.com/
[4] L. Chia-Yi, W. Kuochen, and H. Yi-Ling, "Efficient push-pull based P2P multi-streaming using application level multicast," in Personal Indoor and Mobile Radio Communications (PIMRC), 2010 IEEE 21st International Symposium on, Instanbul 2010, pp. 2586-2590.
[5] Z. Meng, Z. Qian, S. Lifeng, and Y. Shiqiang, "Understanding the Power of Pull-Based Streaming Protocol: Can We Do Better?," Selected Areas in Communications, IEEE Journal on, vol. 25, pp. 1678-1694, 2007.
[6] L. Bracciale, F. Lo Piccolo, D. Luzzi, S. Salsano, G. Bianchi, and N. Blefari-Melazzi, "A push-based scheduling algorithm for large scale P2P live streaming," in Telecommunication Networking Workshop on QoS in Multiservice IP Networks, 2008. IT-NEWS 2008. 4th International, Venice 2008, pp. 1-7.
[7] 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, pp. 1655-1666, 2007.
[8] V. Venkataraman, K. Yoshida, and P. Francis, "Chunkyspread: Heterogeneous Unstructured Tree-Based Peer-to-Peer Multicast," in Network Protocols, 2006. ICNP '06. Proceedings of the 2006 14th IEEE International Conference on, Santa Barbara, CA 2006, pp. 2-11.
[9] Hsin-Wei Wu, "The Design and Implementation of Server Structure for a P2P Video Streaming Delivering System with Guaranteed Latency", Master Dissertation, Hsin-Chu, Taiwan, Department of Computer Science, National Tsing Hua University, 2013.
[10] Zong-Yo Li, "The Design and Implementation of Peer Structure for a P2P Video Streaming Delivering System with Guaranteed Latency", Master Dissertation, Hsin-Chu, Taiwan, Department of Computer Science, National Tsing Hua University, 2013.
[11] Shiu-Shun Peng, "Design and Implementation of Scalable Delivering System for P2P Based Video Streaming Service", Master Dissertation, Hsin-Chu, Taiwan, Department of Computer Science, National Tsing Hua University, 2014.
[12] twitch. [Online] Available: http://www.twitch.tv/
[13] SopCast. [Online] Available: http://www.sopcast.org/
[14] N. Magharei, R. Rejaie, and G. Yang, "Mesh or Multiple-Tree: A Comparative Study of Live P2P Streaming Approaches," in INFOCOM 2007. 26th IEEE International Conference on Computer Communications. IEEE, Anchorage, AK 2007, pp. 1424-1432.
[15] T. Cover and P. Hart, "Nearest neighbor pattern classification," Information Theory, IEEE Transactions on, vol. 13, pp. 21-27, 1967.
[16] C. Gross, B. Richerzhagen, D. Stingl, J. Weber, D. Hausheer, and R. Steinmetz, "GeoSwarm: A multi-source download scheme for peer-to-peer location-based services," in Peer-to-Peer Computing (P2P), 2013 IEEE Thirteenth International Conference on, Trento 2013, pp. 1-10.
[17] D. Chao, J. Yong, X. Shutao, Z. Haitao, and C. Laizhong, "A traffic localization strategy for peer-to-peer live streaming," in Computers and Communications (ISCC), 2013 IEEE Symposium on, Split 2013, pp. 000495-000501.
[18] C. Laizhong and L. Nan, "LANCStream: A Locality-Aware and Network Coding-Based P2P Live Streaming System," in High Performance Computing and Communications & 2013 IEEE International Conference on Embedded and Ubiquitous Computing (HPCC_EUC), 2013 IEEE 10th International Conference on, Zhangjiajie 2013, pp. 496-503.
[19] E. Baralis, A. Bianco, T. Cerquitelli, L. Chiaraviglio, and M. Mellia, "NetCluster: A Clustering-Based Framework for Internet Tomography," in Communications, 2009. ICC '09. IEEE International Conference on, Dresden, Germany, 2009, pp. 1-5.
[20] T. Small, L. Baochun, and B. Liang, "Outreach: peer-to-peer topology construction towards minimized server bandwidth costs," Selected Areas in Communications, IEEE Journal on, vol. 25, pp. 35-45, 2007.
[21] 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, pp. 1446-1456, 2009.
[22] O. Zhipeng, W. Miao, X. Lisong, and B. Ramamurthy, "On providing bounded delay service to subscribers in P2P live streaming systems," in Global Communications Conference (GLOBECOM), 2012 IEEE, Anaheim, CA 2012, pp. 1988-1993.
[23] L. Zimu, W. Chuan, L. Baochun, and Z. Shuqiao, "UUSee: Large-Scale Operational On-Demand Streaming with Random Network Coding," in INFOCOM, 2010 Proceedings IEEE, San Diego, CA 2010, pp. 1-9.
[24] UUSee. [Online] Available:http://www.uusee.com/
[25] G. Kreitz and F. Niemela, "Spotify -- Large Scale, Low Latency, P2P Music-on-Demand Streaming," in Peer-to-Peer Computing (P2P), 2010 IEEE Tenth International Conference on, Delft 2010, pp. 1-10.
[26] M. Goldmann and G. Kreitz, "Measurements on the spotify peer-assisted music-on-demand streaming system," in Peer-to-Peer Computing (P2P), 2011 IEEE International Conference on, Kyoto 2011, pp. 206-211.
[27] Spotify. [Online] Available: https://www.spotify.com/tw/
[28] I. Bermudez, M. Mellia, and M. Meo, "Investigating Overlay Topologies and Dynamics of P2P-TV Systems: The Case of SopCast," Selected Areas in Communications, IEEE Journal on, vol. 29, pp. 1863-1871, 2011.
[29] A. B. Vieira, A. P. C. d. Silva, F. Henrique, G. Goncalves, and P. d. C. Gomes, "SopCast P2P live streaming: live session traces and analysis," presented at the Proceedings of the 4th ACM Multimedia Systems Conference, Oslo, Norway, 2013.
[30] L. Bo, X. Susu, Q. Yang, G. Y. Keung, L. Chuang, L. Jiangchuan, et al., "Inside the New Coolstreaming: Principles, Measurements and Performance Implications," in INFOCOM 2008. The 27th Conference on Computer Communications. IEEE, Phoenix, AZ 2008.
[31] Red5. [Online] Available: https://github.com/Red5