在過去,傳統網路的架構為了擴大網路規模,增加網路的可擴展性,必須透過路由
器或是VLAN技術來切割網段或廣播域,以減少廣播封包對網路的影響,但這種方
式卻降低了IP利用的自由度,以及Host在網路中可移動性,也就是說,在傳統的網路
架構中,網路可擴展性、自由IP的利用和Host的可移動性三者無法兼得,但這些問題
在SDN架構都能獲得解決;而我們也發現,網路中的廣播封包多半屬於ARP協定的封
包,其次為多播封包,由於傳統的交換器無法支援多播的IGMP協定,以至於多播封
包也使用廣播的方式傳送;因此,我們基於SDN技術,提出一個策略,在透過SDN
Controller以及OpenFlow Switch的運作之下,大量減少網路中的廣播和多播封包,而在
我們的模擬結果中證實所提出的架構減少網路中大量的ARP、IGMP和多播封包,除此
之外,我們更是實際佈署在現實的網路中測試,而實際的佈署結果和模擬結果無異,
同樣大幅的降低網路中廣播封包的數量。

In the past, to scale the network or reduce the impact of broadcasts, we cut the network into
subnets by setting routers or divide broadcast domain by setting VLANs, but it makes IP de-
ployment more hard and decrease the network mobility. Fortunately, the software defined net-
work technology can solve the problem. We discover that broadcast occupies most network
bandwidthk, which is caused by Adress Resolution Protocl, and the followed by multicast
caused by L2 switch. Since L2 switch does not support IGMP, it uses broadcast to simulate
multicast behavior. We propose a ”SDN based braodcast reduction” strategy to reduce both the
broadcast and multicast packets. SBBR not only reduces broadcast and multicast packets in the
simulation, but also in the realistic network.

論文審定書i
誌謝ii
摘要iii
Abstract iv
圖次vii
表次ix
Chapter 1 導論1
1.1 簡介. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 研究動機. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 論文貢獻及文章架構. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Chapter 2 研究背景4
2.1 OpenFlow技術. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 RYU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 ARP協定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.4 IGMP協定. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Chapter 3 相關文獻探討11
3.1 傳統網路的問題. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 SDN的解決方法. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Chapter 4 SDN Based Broadcast Reduction 20
4.1 Dispatcher . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
4.2 Ether Handler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
4.3 ARP Handler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
4.4 IGMP Handler . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
Chapter 5 實驗結果28
5.1 ARP的實驗及結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
5.2 多播的實驗及結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
5.3 實際佈署的結果. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Chapter 6 結論以及未來展望44
6.1 結論. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
6.2 未來展望. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Bibliography 45

[1] N. McKeown, Software-defined networking. IEEE INFOCOM Keynote Talk, 2009.
[2] N. McKeown, G. Parulkar, T. Anderson, L. Peterson, H. Balakrishnan, J. Rexford, S. Shenker, and J. Turner, “Openflow: Enabling innovation in campus networks,” ACM SIGCOMM Computer Communication Review, vol. 38, no. 2, pp. 69–74, 2008.
[3] Open Network Fundation [Online] https://www.opennetworking.org/.
[4] Ryu SDN Framework [Online] http://osrg.github.io/ryu/.
[5] W. C. Fenner, “Internet group management protocol, version 2,” RFC 2236, November 1997. http://www.rfc-editor.org/rfc/rfc2236.txt.
[6] Internet Assigned Numbers Authority, [Online] https://www.iana.org/.
[7] D. E. 3rd and J. Abley, “Iana considerations and ietf protocol and documentation usage for ieee 802 parameters,” BCP 141, October 2013.
[8] S. Deering, “Host extensions for ip multicasting,” STD 5, August 1989. http://www.rfc-editor.org/rfc/rfc1112.txt.
[9] E. Bell, A. Smith, P. Langille, A. Rijhsinghani, and K. McCloghrie, “Definitions of managed objects for bridges with traffic classes, multicast filtering and virtual lan extensions,” RFC 2674, August 1999.
[10] A. Myers, T. S. E. Ng, and H. Zhang, “Rethinking the service modle: Scaling ethernet to a million nodes,” ACM SIGCOMM Workshop on Hot Topics in Networking, 2004.
[11] C. Kim, M. Caesar, and J. Rexford, “Floodless in seattle: A scalable ethernet architecture for large enterprises,” ACM SIGCOMM Computer Communication Review, no. 4, pp. 3–14, 2008.
[12] K. Elmeleegy and A. L. Cox, “Etherproxy: Scaling ethernet by suppressing broadcast traffic,” IEEE INFOCOM, pp. 1584–1592, 2009.
[13] C.-H. Chiu and C.-L. Lei, “Etheragent: Scaling ethernet for enterprise and campus networks,” International Journal of Innovative Computing, Information and Control, no. 6, pp. 2465–2483, 2013.
[14] H. Cho, S. Kang, and Y. Lee, “Centralized arp proxy server over sdn controller to cut down arp broadcast in large-scale data center networks,” International Conference on Information Networking, pp. 301–306, 2015.
[15] J. Wang, W. Zhao, S. Yang, J. Liu, T. Huang, and Y. Liu, “Fsdm: Floodless service discovery model based on software-defined network,” IEEE International Conference on Communications Workshops, pp. 230–234, 2013.
[16] P.-W. Chi, Y.-C. Huang, J.-W. Guo, and C.-L. Lei, “Give me a broadcast-free network,” IEEE Global Communications Conference, pp. 1968–1973, 2014.
[17] N. Jehan and A. M.Haneef, “Scalable ethernet architecture using sdn by suppressing broadcast traffic,” International Conference on Advances in Computing and Communications, pp. 24–27, 2015.
[18] K. Kataoka, N. Agarwal, and A. V. Kamath, “Scaling a broadcast domain of ethernet: Extensible transparent filter using sdn,” IEEE International Conference on Computer Communication and Networks, pp. 1–8, 2014.