博碩士論文 103523043 完整後設資料紀錄

DC 欄位 語言
DC.contributor通訊工程學系zh_TW
DC.creator林育瑄zh_TW
DC.creatorYu-Hsuan Linen_US
dc.date.accessioned2016-7-15T07:39:07Z
dc.date.available2016-7-15T07:39:07Z
dc.date.issued2016
dc.identifier.urihttp://ir.lib.ncu.edu.tw:88/thesis/view_etd.asp?URN=103523043
dc.contributor.department通訊工程學系zh_TW
DC.description國立中央大學zh_TW
DC.descriptionNational Central Universityen_US
dc.description.abstract軟體定義網路是近幾年被廣泛討論的網路技術,與傳統交換機相比,軟體定義網路交換機分離了基礎設施層與控制層,集中式的控制器能透過OpenFlow協定控制軟體定義網路交換機與取得子網中軟體定義網路交換機的資訊,讓網路管理人員可依照資訊擺放規則,實現更多創新與彈性的路由方式,且不受限於路由器的品牌或是操控介面,此一特性讓軟體定義網路交換機大量被雲端運算與資料中心所採用,如Google骨幹採用軟體定義網路架構後,大幅提升其頻寬利用率至95%。 行動網路中,物聯網需求快速成長,3GPP Release 12中除了規範更低的頻寬使用、支援半雙工與PSM的Cat. 0,在大量的裝置信令與小資料封包的情況下,如何有效處理這些封包,目前仍然沒有最佳解決方案。若將軟體定義網路與網路功能虛擬化結合,並且應用於核心網路,則在雲端叢集上能夠有快速部屬與提供細膩規則的優勢,因此被認為是解決此問題的方法之一。 軟體定義網路交換機中的流項提供很大的彈性,OpenFlow v1.5中包含match field等七個欄位。而硬體三態內容尋址記憶體是提供高速比對的重要角色,但由於三態內容尋址記憶體的昂貴與耗能,在規則的儲存上會有上限,若未經規劃,大型網路中會出現規則數超過軟體定義網路交換機容量或流量集中某一節點的情況,因此在本論文中,提出規則放置策略,在路由基礎規則配置後,考慮無效頻寬最小化動態規則空間使用。 本論文分別在EPC、ITALYNET、Cloud拓樸下進行模擬,分析其TCAM使用與頻寬利用,我們所提出的方法,在TCAM的使用上非常接近TCAM最小化的選擇,在頻寬使用率則有3~28%的提升。zh_TW
dc.description.abstractSoftware-defined networking is a network technology which has been widely discussed in recent years. Comparison to the traditional switch, the SDN switch separate the infrastructure layer and control layer. Centralized controller can manage the SDN switch via OpenFlow protocol. By following the protocol controller can easily get the information of the subnet which made the network management can place the rules to achieve more innovative and flexible routing, and most important of all the switch is not limited by the brand or the control interface. This feature allows cloud computing and datacenter using SDN switch to improve their network utility. For instance, Google backbone with SDN architecture which significantly increase its bandwidth utilization to 95%. In mobile network, the internet-of-things devices increase rapidly. The 3GPP Release 12 specifications support lower bandwidth usage, half-duplex and the PSM Cat. 0. However in the case of the large number of small data packets and signaling message still have not a good method to deal it. Combination of the SDN and NFV using in the core network can provide fast and fine grained management which is considered one of the solution of this problem. The flow entry in OpenFlow protocol version 1.5 contains seven fields. And the TCAM is a hardware architecture to provide high speed matching. Because the TCAM is expensive and power hungry, the rule space which would be limited. If we did not well place the rules in the large scale network, the network traffic or the rule capacity may concentrate in one node which make the network load and rule capacity unbalance. So in this paper, we proposed a rule placement method which is minimizing TCAM usage then consider the effective bandwidth utilization. In this paper, using the EPC, ITALYNET and Cloud topology to simulate and analyze the TCAM usage and bandwidth utilization. In our proposed method, the usage of the TCAM is quite close to minimizing TCAM method, and the bandwidth utilization rates have 3 to 28 % improvement.en_US
DC.subject軟體定義網路zh_TW
DC.subject規則放置zh_TW
DC.subjectTCAM空間zh_TW
DC.subject頻寬使用zh_TW
DC.title軟體定義網路中減少TCAM空間使用及頻寬有效利用之規則放置方法研究zh_TW
dc.language.isozh-TWzh-TW
DC.titleStudy of Rule Placement Scheme for Minimizing TCAM Space and Effective Bandwidth Utilization in Software-Defined Networkingen_US
DC.type博碩士論文zh_TW
DC.typethesisen_US
DC.publisherNational Central Universityen_US

若有論文相關問題,請聯絡國立中央大學圖書館推廣服務組 TEL:(03)422-7151轉57407,或E-mail聯絡  - 隱私權政策聲明