臺灣博碩士論文加值系統

在大數據的時代中，多重Domain的架構逐漸被校園以及企業網路所採用，在龐大資料的流動下，網路建設與管理必須更加靈活且具有擴展性，而傳統網路由於硬體設施的限制而慢慢不敷使用，SDN技術的發展，將網路分成控制平面及數據平面，使得管理人員可以從集中式控制平面動態來規劃數據平面，使其能更加彈性解決網路中所發生的問題。
在本論文當中，我們於多重Domain的架構下利用SDN進行協同式異質流管理，由於同個時間內每個Domain的負載狀況盡不相同，我們可以將負載大的Domain之流量導入負載較輕的Domain以進行協同式傳輸，如此一來，負載過大的Domain能有更多的路徑選擇，並藉此讓其他Domain分擔其負載，以提升整體網路的Throughput。
　　在協同式傳輸的過程當中，協助Domain方的網路負載狀況不會持續相同，當其負載量變大時，外部資料流與內部資料流將會競爭頻寬，可能導致協助方自身的資料流無法獲得傳輸保障，有鑒於此，我們將異質流的觀念導入，利用權重來區分協同式傳輸以及自身Domain的資料流，在不同權重的規劃下，讓高權重資料流可享有較多的的頻寬以保障其服務品質，而低權重的資料流也有一定的頻寬傳輸以避免飢餓的情況發生，透過模擬實驗證明，我們所提出的協同式異質流管理方法(Cooperative Management of Heterogeneous Flows，即CMHF)可以有效提升整體網路的Throughput，同時也能保障Domain自身資料流的傳輸品質。

In the era of big data, the multi-domain architecture is gradually being applied to campus and enterprise networks. With the massive increase in traffic flows, network architecture and management must be more flexible and expandable. Traditional networks are constrained by their hardware design, which may not be able to cope with the quality demand of traffics. The technology of software-definded networking(SDN) decouples a network into control and data planes. In this way, network administrators can easity direct the data plane through a centralized control plane, which can flexibly deal with problems occurred in the network.
In this thesis, we adopt SDN to realize cooperative heterogeneous flow management in a multi-domain network. Because the amount of load in each domain is different, we can redirect some flows in a busy domain to other domains through cooperative routing. This provides more choices of paths for that domain and distributes its workload to other domains, so as to improve the overall performance.
During the process of cooperative rounting, the load of an assisting domain may not always be stable. When its work load becomes heavy, the exterior flows and interior flows will compete the bandwidth, which may not support QoS for interior flows. In view of this, we propose the concept of heterogeneous flows by giving them different priorities. In particular, high- priority flows can consume more bandwidth for Qos support. On the other hand, low-priority flows still have some bandwidth to use to avoid them starvation. Through simulations, we show that our proposed cooperative routing mechanism can raise netowork performance and ensure that each domain can support QoS for their interior flows.

目錄
論文審定書 i
誌謝 ii
摘要 iii
Abstract iv
目錄 vi
圖次 viii
表次 x
第一章 緒論 1
1.1 簡介 1
1.2 研究動機 3
1.3 論文架構與貢獻 4
第二章 研究背景 5
2.1 SDN網路架構 5
2.2 OpenFlow協定 7
2.2.1 Flow Table 8
2.2.3 Group Table 10
2.2.4 Meter Table 12
2.3 常見的SDN Controller介紹 13
第三章 相關文獻探討 16
3.1 分散式架構下Controller通訊 17
3.2 分散式Controller架構之負載平衡 18
3.3 Link負載平衡 18
第四章 問題定義 20
第五章 研究方法 21
5.1 網路建置與Controller通訊 23
5.2 Switch資訊蒐集 25
5.3 網路壅塞監控 28
5.4 負載平衡處理 29
5.4.1 同質流負載平衡機制 30
5.4.2 異質流負載平衡機制 33
第六章 實驗分析與討論 37
6.1 實驗工具與模擬環境 37
6.2 網路拓樸與流量設定 38
6.3 實驗(一)：網路整體傳輸效能 39
6.4 實驗(二)：協同式傳輸效能 46
6.4.1 Throughput效能比較 47
6.4.2 Packet loss rate分析 48
第七章 結論與未來展望 50
7.1 結論 50
7.2 未來展望 50
參考文獻 51
附錄A Openvswitch修改 55

[1]International Organization for Standardization. [Online]. Available: https://www.iso.org/home.html.
[2]N. McKeown, “Software-defined networking,” IEEE INFOCOM Keynote Talks, April, 2009.
[3]N. McKeown, T. Anderson, H. Balakrishnan, G. Parulkar, L. Peterson et al., ‘‘OpenFlow: Enabling innovation in campus networks,’’ SIGCOMM Computer. Communication. Review., vol. 38, no. 2, pp. 69–74, 2008.
[4]S. K. Dey and M. M. Rahman, “Flow based anomaly detection in software defined networking: A deep learning approach with feature selection method,’’ International Conference on Electrical Engineering and Information & Communication Technology, pp. 630-635, 2018.
[5]M. Robinson, M. Milosavljevic, P. Kourtessis, S. Fisher, G. P. Stafford et al., “QoE based holistic traffic engineering in SDN enabled heterogeneous transport networks,’’ International Conference on Transparent Optical Networks, pp. 1-4, 2017.
[6]S. Tomovic, N. Lekic, I. Radusinovic,and G. Gardasevic, “A new approach to dynamic routing in SDN networks,’’ Mediterranean Electrotechnical Conference, pp. 1-6, 2016.
[7]ONF. [Online]. Available: https://opennetworking.org/
[8]OpenFlow 1.0. [Online]. Available: https://opennetworking.org/wp-content/uploads/2013/04/openflow-spec-v1.0.0.pdf
[9]OpenFlow 1.1. [Online]. Available: https://opennetworking.org/wp-content/uploads/2014/10/openflow-spec-v1.1.0.pdf
[10]OpenFlow 1.3. [Online]. https://opennetworking.org/wp-content/uploads/2014/10/openflow-spec-v1.3.0.pdf
[11]NOX. [Online]. Available: https://github.com/noxrepo/nox
[12]POX. [Online]. Available: https://noxrepo.github.io/pox-doc/html/
[13]OpenDaylight. [Online]. Available:https://www.opendaylight.org/
[14]ONOS. [Online]. Available:https://onosproject.org/
[15]Ryu SDN Framework. [Online]. Available: https://osrg.github.io/ryu-book/zh_tw/Ryubook.pdf
[16]L. Li and Q. Xu, “Load balancing researches in SDN: A survey,’’ IEEE International Conference on Electronics Information and Emergency Communication, pp. 403-408, 2017.
[17]H. G. Ahmed and R. Ramalakshmi, “Performance analysis of centralized and distributed SDN controllers for load balancing application,’’ International Conference on Trends in Electronics and Informatics, pp. 758-764, 2018.
[18]A. Tootoonchian and Y. Ganjali, “Hyperflow: A distributed control plane for openflow, ’’ Internet Network Management Workshop, pp. 3-3, 2010.
[19]T. Koponen, M. Casado, N. Gude, J. Stribling, L. Poutievski et al. “ONIX: A distributed control platform for large-scale production networks,’’ USENIX Conference. Operating Systems Design and Implementation., pp. 1-6, 2010.
[20]J. Stribling, Y. Sovran, I. Zhang, X. Pretzer, J. Li et al. “Flexible, wide-area storage for distributed systems with wheelfs,’’ USENIX Symposium on Networked Systems Design and Implementation. vol. 9, pp. 43-58 , 2009
[21]P. Lin, J. Bi, and Y. Wang. ’’East-west bridge for SDN network peering,’’ Frontiers in Internet Technologies. Springer, Berlin, Heidelberg. pp. 170-181, 2013.
[22]W. H. F. Aly, ’’Controller adaptive load balancing for SDN networks,’’ Eleventh International Conference on Ubiquitous and Future Networks, pp. 514-519, 2019.
[23]P. T. Duy, D. T. T. Hien, H. P. Qui and V. Pham, ’’Aloba: A mechanism of adaptive load balancing and failure recovery in distributed SDN controllers, ’’ IEEE International Conference on Communication Technology, pp. 1322-1326, 2019.
[24]T. Hu, P. Yi, J. Zhang,and J. Lan, ’’A distributed decision mechanism for controller load balancing based on switch migration in SDN,’’ China Communications, vol. 15, no. 10, pp. 129-142, 2018.
[25]T. Oliveira ,and S. Sargento, ’’QoE-based load balancing of OTT video content in SDN networks,’’ IEEE Symposium on Computers and Communications, pp. 1-6, 2019.
[26]S. Patil, ’’Load balancing approach for finding best path in SDN,’’ International Conference on Inventive Research in Computing Applications, pp. 612-616, 2018.
[27]F. S. Fizi and S. Askar, ’’A novel load balancing algorithm for software defined network based datacenters,’’ International Conference on Broadband Communications for Next Generation Networks and Multimedia Applications , pp. 1-6, 2016.
[28]X. Liu and Y. Xiang, ’’Disturbance based dynamic load balancing routing mechanism under SDN,’’ International Conference on Computer Science and Network Technology , pp. 651-656, 2016.
[29]T. T. Huong, N. D. D. Khoa, N. X. Dung ,and N. H. Thanh, ’’A global multipath load-balanced routing algorithm based on reinforcement learning in SDN,’’ International Conference on Information and Communication Technology Convergence, pp. 1336-1341, 2019.
[30]U. Mahlab, Y. Wolbrum and Z. Erlichson, ’’Traffic load-balancing for software-defined elastic optical networks based cross-entropy technique,’’ International Conference on Transparent Optical Networks, pp. 1-4, 2019.
[31]Y. -C. Wang and E. -J. Chang, ’’Cooperative Flow Management in Multi-domain SDN-based Networks with Multiple Controllers, ’’ IEEE International Conference on Smart Communities: Improving Quality of Life Using ICT, IoT and AI, pp. 82-86, 2020.
[32]XPub / XSub. [Online]. Available: https://netmq.readthedocs.io/en/latest/xpub-xsub/
[33]Networkx. [Online]. Available:https://networkx.github.io/
[34]OpenvSwitch. [Online]. Available:https://www.openvswitch.org/
[35]Mininet. [Online]. Available: http://mininet.org/
[36]Iperf. [Online]. https://iperf.fr/
[37]I. Stoica, H. Zhang and T. S. E. Ng, ’’A hierarchical fair service curve algorithm for link-sharing, real-time, and priority services, ’’ IEEE/ACM Transactions on Networking, vol. 8, no. 2, pp. 185-199, 2000
[38]Stochastic Switching. [Online]. Available:https://github.com/saeenali/openvswitch/wiki/Stochastic-Switching-using-Open-vSwitch-in-Mininet