在大型資料的網路環境中，既有的分散式網路架構管理與頻寬使用效率隨著拓樸的規模增大而顯得更加成效不彰，為了提升網路效率，簡化管理模式，Google、Yahoo、Facebook等多家指標型的大企業投入了新一代的軟體定義式網路SDN(Software Defined Networking)的架構與技術發展。

SDN 將網路集中管理，並提供可程式化的機制，對於企業或網路資料中心來說，可以藉此獲得自動處理與動態因應變化的好處，為網路資源的調度提供靈活度與彈性，但為此目的，SDN交換機需要設置更快更昂貴的TCAM，來應對存取頻繁，資料更新迅速的SDN流表(flow table)，且為了節省交換機成本，TCAM的容量往往有其上限，若達使用上限時將自動移除優先度較低的流向(flow entry) ，但這往往無法有效率的使用TCAM中的規則空間(rule space)，本論文探討利用ROHC壓縮封包時的封包標頭合併傳送與解壓縮機制，使得企業網路各區網間的SDN switch達到有效減少flow table中的rule數量，進而增進TCAM使用效率，在不提高switch記憶體成本的情況下，維持路由的正常工作，提高flow entry的hit rate，更可降低區域網路間的流量。

The distributed networks will be lower performance and harder to management when the network topology expansion in the public network. The iconic network businesses: Google、Yahoo and Facebook start into research and development the SDN(Software Defined Networking) for improve the network efficiency and simplify the network management system. Network intelligence is centralized in SDN that maintain the network, the SDN also provide the programmatically configured. The SDN for the enterprises and network data centers can get the benefit of the network automated and dynamically adjust network-wide traffic flow to meet changing needs. The SDN switch need the faster but more expensive TCAM(Ternary Content Addressable Memory) for the SDN flow table or refresh the network data and access data constantly. In order to lower the SDN switch cost, it always need to limit the TCAM size of the switch, when the flow entries exceed the TCAM maximum size, the SDN switch must need to remove the lower priority flow entry automatically, it will let the SDN TCAM’s rule space get lower efficiency. This thesis is using the ROHC to compress the network packet’s head for reduce the TCAM size usage. The ROHC will improve the TCAM efficiency but no need increase the SDN switch cost base on the same TCAM size, it still can let network work stable, higher flow entry rate and reduce the network traffic.

[1]. 網路資料：https://www.ithome.com.tw/node/77353
[2]. 網路資料：https://www.ithome.com.tw/tech/87693
[3]. 網路資料：https://www.sdxcentral.com/sdn/definitions/inside-sdn-architecture/
[4]. http://www.informit.com/articles/article.aspx?p=2451956&seqNum=2 "Foundations of Modern Networking: Background and Motivation of Software-Defined Networks (SDN) By William Stallings", Dec 3, 2015
[5] “OpenFlow Switch Specification. Version 1.4.0 (Wire Protocol 0x05)”, October 14, 2013
[6]. 網路資料：https://www.netadmin.com.tw/article_content.aspx?sn=1610070003
[7]. “OpenFlow Switch Specification Version 1.4.0”, Open Networking Foundation, Oct 14, 2013
[8]. Subhasis Banerjee and Kalapriya Kannan, "Tag-In-Tag: Efficient flow table management in SDN switches", Proc. Of 2014 10th CNSM, pp. 17-21, Nov. 2014
[9]. Kun Qiu, Jing Yuan, Jin Zhao, Xin Wang, Stefano Secci and Xiaoming Fu, "Fast Lookup Is Not Enough: Towards Efficient and Scalable
Flow Entry Updates for TCAM-based OpenFlow Switches", Proc. Of 2018 IEEE 38th ICDCS , pp. 2-6, July 2018
[10]. RObust Header Compression (ROHC): Framework and four profiles: RTP, UDP, ESP, and uncompressed , IETF RFC 3095, 2001.
[11]. 網路資料：http://www.effnet.com/products/ROHC-v2/
[12]. Li Li-fu, Li Hai-wen, Li Hong-liang, Gu Yong-jun, “Research of ROHC for PoC Service Based on 3G Cellular Network”,
2017 IEEE 2nd Advanced Information Technology, Electronic and Automation Control Conference (IAEAC), pp. 900-904, Mar. 2017.
[13]. 厲群、王春曉，「ROHC協議分析與建模」，計算机工程与設計, CSCD 北大核心 2008年第13期 3309-3312
[14] 網路資料：http://www.sharetechnote.com/html/Handbook_LTE_ROHC.html
[15]. 網路資料：https://slideplayer.com/slide/7844260/
[16]. 網路資料： https://www.slideshare.net/kentaroebisawa/fpgax-open-flowfpga20150201english
[17]. Open vSwitch：https://www.openvswitch.org/
[18] 網路資料：https://www.rightscale.com/blog/cloud-management-best-practices/docker-vs-vms-combining-both-cloud-portability-nirvana