近十年以來，隨著IP網路科技的持續演進，各式新興的網路程式也隨之發展，其中不乏許多互動式應用如網路電話、支援影音串流的視訊會議等，皆扮演科技應用的重要腳色。然而這些具備高頻寬或低延遲特性的即時服務在面臨嚴重的網路壅塞時，在傳輸效能或是使用者經驗上都會造成莫大影響。面對此重大議題，過去幾年傳統的IP網路提供Integrated Service(IntServ)和Differentiated Service (DifSer)兩大方式，試圖在現有的網路架構上提供服務品質，卻因網路硬體設備限制和大規模網路所衍生的複雜性而未顯成效。
近幾年軟體定義網路的興起，顛覆了傳統網路的運作方式，其透過控制層和資料層的分離，使得彈性化、客製化的網路監控、路由和管理能得以實現。在本篇論文中，我們基於軟體定義網路的架構，實作了一服務品質管理系統並提出一個動態路徑演算法，其旨在透過即時監控數據和模型，動態產生一最低負載成本之路徑。另外，透過整合一個基於機器學習演算法的流量辨識服務平台，我們設計一種頻寬管理機制，其能根據應用程式辨識結果，幫助管理機制在網路壅塞時，針對現有運行流量，依照使用者所設定之應用程式優先次序進行不同的速率限制。測試數據顯示我們所提出的動態路徑演算法在實體網路以及模擬網路環境中能在耗費低時間成本的前提下，在檔案傳輸以及影音串流服務提供良好的服務品質。另外，我們也設計了數個使用情境測試提出的頻寬管理機制，其結果顯示，針對不同的網路應用程式，系統皆能在一秒至數秒內完成流量辨識並根據優先次序高低給予特定流量速率限制。

In previous decades, along with the continuous evolution of the Internet Protocol-based network, numerous Internet programs that have played critical roles in the current technological era are being developed inclusive of interactive applications like VoIP, video conferencing with audio, and video streaming. Nevertheless, the transmission performance and user experience of these instant services that require high bandwidth or low latency may suffer when serious traffic congestion occurs. In approaching this issue, two mechanisms – Integrated Service and Differentiated Service – were proposed in traditional IP networks to provide quality of service (QoS) with the existing network architecture; however, results have been unsuccessful due to limitations in hardware devices and complexities from the scalability.
Software-defined networking in recent years has emerged, where the decoupling of the control plane and data plane altering the original operations makes it possible to provide flexible networking monitoring, routing, and management. In this paper, we implement a QoS management system with a dynamic routing algorithm based on the software-defined networking architecture. It aims to generate an optimal route with minimum cost based on real-time monitored statistics and the cost model. Moreover, not only is it integrated with a flow classification engine based on machine learning algorithms, but the system is also equipped with a bandwidth management mechanism that can enforce rate-limit policies for each application with different priority under traffic congestion. Experimental results show that the proposed algorithm can provide excellent QoS with little overhead for both file transferring and streaming services. In addition, several test cases were designed to verify the bandwidth management mechanism, where the result shows that QoS policies can be fulfilled in one to several seconds for different applications.

Chapter 1 Introduction 1
Chapter 2 Related Works 4
2.1 Software-Defined Networking 4
2.2 OpenFlow 5
2.3 Flow Classification Engine 7
2.4 Related Works of OpenFlow-enabled QoS Systems 8
Chapter 3 System Design and Implementation 13
3.1 System Overview 14
3.1.1 Floodlight Controller 14
3.1.2 Flow Classifier 15
3.1.3 Authentication Server 16
3.2 System Implementation 16
3.2.1 Monitor Service 16
3.2.2 Evaluator Service and Event Database 17
3.2.3 Classified Flow Database and History Flow Database 18
3.2.4 Event Handler 19
3.2.5 Forwarding Module 21
3.2.6 QoS Manager GUI 24
3.2.7 Authentication Server 26
Chapter 4 Experimental Results 30
4.1 Route Calculation 31
4.2 HTTP File Transferring 35
4.3 RTSP UDP Streaming 39
4.4 Reaction Time 43
4.5 Rate-limit 45
Chapter 5 Conclusion and Future Works 49 

[1] R. Braden, D. Clark, and S. Shenker, “Integrated services in the internet architecture: an Overview,” RFC 1633, Internet Engineering Task Force, June 1994.
[2] R. Braden, L. Zhang, S. Berson, S. Herzog, and S. Jamin, “Resource reservation protocol (rsvp) – version 1 functional specification,” RFC 2205, Internet Engineering Task Force, September 1997.
[3] S. Blake, M. Carlson, Z. Wang, W. Weiss, “An Architecture for Differentiated Services”, RFC2475, Internet Engineering Task Force, December 1998.
[4] Floodlight. [Online]. Available: http://floodlight.openflowhub.org
[5] OpenFlow. [Online]. Available: http://archive.openflow.org
[6] Open Networking Foundation. [Online]. Available: http://opennetworking.org
[7] Chia-Chin Chou, “A Cloud based Application Classification Service Platform with Machine Learning Algorithms,” M.S. thesis, Inst. Com. Eng., National Tsing Hua Univ., Hsinchu, Taiwan, 2013.
[8] N.-F. Huang, G.-Y. Jai, C.-H Chen, H.-C. Chao, “On the Cloud-Based Network Traffic Classification and Applications Identification Services,” International Conference on Selected Topics in Mobile and Wireless Networking (iCOST), 2012, pp. 36-41.
[9] N.-F. Huang, G.-Y. Jai, H.-C. Chao, Y.-J. Tzang, H.-Y Chang, “Application traffic classification at the early stage by characterizing application rounds,” Information Sciences (SCI), Vol. 232, May 2013, pp. 130-142.
[10] Bari, M.F. ; Chowdhury, S.R. ; Ahmed, R. ; Boutaba, R., “PolicyCop: An Autonomic QoS Policy Enforcement Framework for Software Defined Networks,” IEEE SDN for Future Networks and Services (SDN4FNS), 2013.
[11] Hilmi E Egilmez, S Tahsin Dane, K Tolga Bagci, and A Murat Tekalp, “OpenQoS: An openflow controller design for multimedia delivery with end-to-end quality of service over software-defined networks”, in Signal & Information Processing Association Annual Summit and Conference (APSIPA ASC), 2012 Asia-Pacific, pages 1–8, 2012.
[12] Lara, A., Kolasani, A., Ramamurthy, B., “Simplifying network management using Software Defined Networking and OpenFlow,” IEEE International Conference on Advanced Networks and Telecommuncations Systems (ANTS), 2012.
[13] Open vSwitch, An Open Virtual Switch. [Online]. Available: http://openvswitch.org/
[14] Iris Bueno, José Ignacio Aznar, Eduard Escalona, Jordi Ferrer, Joan Antoni García-Espín, “An OpenNaaS based SDN Framework for Dynamic QoS Control,” IEEE SDN for Future Networks and Services (SDN4FNS), 2013.
[15] S. R. Chowdhury, M. F. Bari, R. Ahmed, and R. Boutaba, “PayLess: A Low Cost Netowrk Monitoring Framework for Software Defined Networks,” in 14th IEEE/IFIP Network Operations and Management Symposium, 2014.
[16] A. D. Ferguson et al., “Participatory networking: an API for application control of SDNs,” in Proceedings of the ACM SIGCOMM, 2013, pp. 327–338.
[17] H. E. Egilmez, S. Civanlar, and A. M. Tekalp, “An Optimization Framework for QoS-enabled Adaptive Video Streaming over OpenFlow Networks,” IEEE Trans. on Multimedia, 2013.
[18] H. E. Egilmez, B. Gorkemli, A. M. Tekalp, and S. Civanlar, “Scalable video streaming over OpenFlow networks: an optimization framework for QoS routing,” in Proc. IEEE International Conference on Image Processing (ICIP), Sept. 2011, pp. 2241–2244.
[19] S. Cabuk, C. I. Dalton, K. Eriksson, D. Kuhlmann, H. V. Ramasamy, G. Ramunno, A.-R. Sadeghi, M. Schunter, and C. St ¨uble, “Towards Automated Security Policy Enforcement in Multi-tenant Virtual Data Centers,” Journal of Computer Security, 18(1):89–121, 2010.
[20] W. Kim, P. Sharma, J. Lee, S. Banerjee, J. Tourrilhes, S.J. Lee, and P. Yalagandula, “Automated and scalable qos control for network convergence”, Proc. INM/WREN, 2010.
[21] Nick McKeown et. al., “Openflow: enabling innovation in campus networks”, SIGCOMM Comput. Commun. Rev., 38(2):69–74, 2008.
[22] Z. Wang and J. Crowcroft. “Quality-of-service routing for supporting multimedia applications,” IEEE Journal on Selected Areas in Communications, 14(7):1228–1234, 1996.
[23] OpenFlow Switch Specification, Version 1.0.0 ( Wire Protocol 0x01 ). [Online]. Available: http://www.openflow.org/documents/openflow-spec-v1.0.0.pdf
[24] Internet Engineering Task Force (IETF). [Online]. Available: http://www.ietf.org/
[25] Mininet. [Online]. Available: http://mininet.org/
[26] VLC. [Online]. Available: http://www.videolan.org/vlc/
[27] MSU Video Quality Measurement Tool. [Online]. Available: http://compression.ru/video/quality_measure/video_measurement_tool_en.html
[28] OpenDayLight. [Online]. Available: http://www.opendaylight.org/
[29] Ryu. [Online]. Available: http://osrg.github.io/ryu/
[30] OpenFlow Switch Specification, Version 1.3.0 ( Wire Protocol 0x04). [Online]. Available: https://www.opennetworking.org/images/stories/downloads/specification/openflow-spec-v1.3.0.pdf