臺灣博碩士論文加值系統

在此篇論文中我們提出了一種在軟體定義網路中一基於統計的動態繞送路徑分配機制，在該機制下，我們利用 OpenFlow 中的群組表 (Group table) 使得網路中的資料流 (traffic) 能夠不再只是被指派到某條調特定的繞送路徑 (routing path) 上，而是能夠將一條資料流分流到多條不同路徑上，最後在到達目地端，而在決定要分流該資料流中多少比例到其他路徑上時，我們會依據最好的路徑 (shortest-path) 中還剩多少頻寬可以使用，若該路徑的頻寬不夠時，再尋找次好的繞送路徑看該路徑有多少頻寬可使用，不斷依續尋找，直到該資料流被分流完
為止，這樣一來可避免在網路資源夠的前提下仍然造成網路壅塞的情況，而網路壅塞確實是造成網路服務品質下降的主因，根據我們研究過的文獻中，在解決網路壅塞的情況時，不外乎像是對某些資料流限流、繞送路徑的選擇與調配... 等等。
我們利用 mininet 來模擬 SDN 網路，在控制器 (Controller) 上執行我們提出的演算法來驗證我們所提出的方法相對於以往有哪些提升，在實驗中我們考慮四種類型的資料流，分別為盡最大努力 (best-effort)、檔案傳輸、串流媒體以及網路電話等四種，實驗結果顯示，對於盡最大努力類型的資料流，透過我們的方法能夠提升網路資源使用率，進而提升該類型資料流使用者的傳輸吞吐量以及降低風包遺失率 (packet loss rate)；而對於需保證送達 (TCP) 類型的資料流，像是檔案傳輸，透過我們的方法能透提升傳輸時的速率，另一方面也就是減少傳輸時間；再來對於串流媒體類停的資料流，透過我們的方法能夠避免在媒體在串流中，某些時候資料量突然急遽上升所造成的失真；最後，對於像是網路電話這種對延遲敏感度相當高的應用，由於透過我們的方法能夠避免在網路資源夠的前提下仍然造成網路壅塞的情況，因此進而降低該類型資料流在傳輸過程中的來回延遲(Round Trip Time, RTT)。

In this thesis, we proposed a Dynamic Separating Routing Mechanism in Software Defined Network, in this mechanism, we leverage the ”group table” defined in OpenFlow protocol to divide single traffic in to multiple part, then assign multiple routing path for these part rather than assign single routing path for single traffic. In order to determine how to separate this traffic, the mechanism put the traffic into the best available routing path, once the residual bandwidth of this routing path is not enough for the traffic, calculate next-best available path and get residual bandwidth of the next-best vailable path, and so on, until the traffic can be shared into these residual bandwidth. Dynamic Separating can avoid the scenario which got congestion link on the premise that the network resource is enough, the congestion is always the key to influence the performance on networks, according to our researched papers, to solve congestion event in the networks such like limit some traffic in the network, routing path assignment and reallocation, etc.
We use mininet to simulate SDN environment and running our proposed algorithm on RYU controller to evaluate how our method improve compare to the traditional way, in the experiment, we consider four type of network traffic, they are best-effort, file transfer, video streaming and VoIP, the experiment result shows that for best-effort traffic, our method can improve network resource utilization to get higher throughput, also reduce the packet loss rate; for file transfer application, our method can accelerate data rate during the file transfer, on another word, it can saving transmission time; for video streaming application, our method can avoid distortion caused by peak bit rate of streaming; finally, for VoIP traffic, our method demonstrate shorter round trip delay and better user satisfaction compare with traditional way.

Acknowledgments i
Abstract iii
List of Figures ix
List of Tables xi
Chapter 1 Introduction 1
1.1 Different Between SDN and Traditional Network . . . . . . . . . . . . 1
1.2 SDN Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 OpenFlow Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Flow Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.2 Group Table . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Chapter 2 Related Work 9
Chapter 3 System Model and Problem Formulation 12
3.1 System Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2 Problem Formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.1 Network Resource Waste . . . . . . . . . . . . . . . . . . . . . 14
3.2.2 Resource Allocation . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.3 Optimum Routing Path Assignment . . . . . . . . . . . . . . 16
Chapter 4 Proposed Algorithms 17
4.1 Traffic Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.2 Available Path Discovery . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.3 Dynamic Separating Algorithm . . . . . . . . . . . . . . . . . . . . . 24
4.4 Traffic Handler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
4.5 Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.6 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Chapter 5 Experiments 37
5.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.2 Computation Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
5.3 Best-effort and File Transfer Applications . . . . . . . . . . . . . . . 42
5.3.1 Routing With Traditional Way . . . . . . . . . . . . . . . . . 43
5.3.2 Routing With Dynamic Separating . . . . . . . . . . . . . . . 48
5.3.3 Comparison of Best-effort And File Transfer . . . . . . . . . . 54
5.4 Streaming Video and VoIP Applications . . . . . . . . . . . . . . . . 61
5.4.1 Routing With Traditional Way . . . . . . . . . . . . . . . . . 63
5.4.2 Routing With Dynamic Separating . . . . . . . . . . . . . . . 67
5.4.3 Comparison of Streaming Video and VoIP . . . . . . . . . . . 70
Chapter 6 Conclusion 77
Bibliography 79

[1] What are sdn controllers (or sdn controllers platforms)? https://www.sdxcentral.com/sdn/definitions/sdn-controllers/, . [Online].
[2] Software-defined networking (sdn) definition. https://www.opennetworking.org/sdn-resources/sdn-definition, . [Online].
[3] Openflow switch specification. https://www.opennetworking.org/images/stories/downloads/sdn-resources/onf-specifications/openflow/openflow-spec-v1.4.0.pdf.
[4] Diego Kreutz, Fernando MV Ramos, P Esteves Verissimo, C Esteve Rothenberg, Siamak Azodolmolky, and Steve Uhlig. Software-defined networking: A comprehensive survey. Proceedings of the IEEE, 103(1):14–76, 2015.
[5] The zettabyte era —trends and analysis. http://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/vni-hyperconnectivity-wp.html, 2016. [online].
[6] Davide Adami, Gianni Antichi, Rosario G Garroppo, Stefano Giordano, and Andrew W Moore. Towards an sdn network control application for differentiated traffic routing. In 2015 IEEE International Conference on Communications (ICC), pages 5827–5832. IEEE, 2015.
[7] Mininet. http://mininet.org/. [Online].
[8] Ryu sdn controller. https://osrg.github.io/ryu/. [Online].
[9] Hailong Zhang and Jinyao Yan. Performance of sdn routing in comparison with legacy routing protocols. In Cyber-Enabled Distributed Computing and Knowledge Discovery (CyberC), 2015 International Conference on, pages 491–494. IEEE, 2015.
[10] Mun Choon Chan and Yow-Jian Lin. Behaviors and effectiveness of rerouting: a study. In IEEE International Conference on Communications, 2005. ICC 2005. 2005, volume 1, pages 218–223. IEEE, 2005.
[11] Yong Liu and AL Narasimha Reddy. A fast rerouting scheme for ospf/is-is networks. In Computer Communications and Networks, 2004. ICCCN 2004. Proceedings. 13th International Conference on, pages 47–52. IEEE, 2004.
[12] Insun Jang, Sukjin Choo, Myeongsu Kim, Sangheon Pack, and Myung-Ki Shin. Optimal network resource utilization in service function chaining. In 2016 IEEE NetSoft Conference and Workshops (NetSoft), pages 11–14. IEEE, 2016.
[13] Network function virtualization. https://www.ericsson.com/res/docs/whitepapers/network-functions-virtualization-and-software-management.pdf. [Online].
[14] Sai Qian Zhang, Qi Zhang, Hadi Bannazadeh, and Alberto Leon-Garcia. Network function virtualization enabled multicast routing on sdn. In 2015 IEEE International Conference on Communications (ICC), pages 5595–5601. IEEE, 2015.
[15] Quality of service. http://www.voip-info.org/wiki/view/QoS. [Online].
[16] Hung-Lin Huang, Min-Huang Ho, Shao-Jui Chen, Yi-Feng Chen, and Wei-Jen Wang. A dynamic network bandwidth allocation mechanism under network qos constraints. In 2015 IEEE International Conference on Smart City/SocialCom/SustainCom (SmartCity), pages 963–968. IEEE, 2015.
[17] Anand V Akella and Kaiqi Xiong. Quality of service (qos)-guaranteed network resource allocation via software defined networking (sdn). In Dependable, Autonomic and Secure Computing (DASC), 2014 IEEE 12th International Conference on, pages 7–13. IEEE, 2014.
[18] Jason Min Wang, Ying Wang, Xiangming Dai, and Brahim Bensaou. Sdnbased multi-class qos-guaranteed inter-data center traffic management. In Cloud Networking (CloudNet), 2014 IEEE 3rd International Conference on, pages 401–406. IEEE, 2014.
[19] Alain Hore and Djemel Ziou. Image quality metrics: Psnr vs. ssim. In Pattern recognition (icpr), 2010 20th international conference on, pages 2366–2369. IEEE, 2010.
[20] Mohammed Rana Basheer and Sarangapani Jagannathan. R-factor: a new
parameter to enhance location accuracy in rssi based real-time location systems. In 2009 6th Annual IEEE Communications Society Conference on Sensor, Mesh and Ad Hoc Communications and Networks, pages 1–9. IEEE, 2009.