Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/7649
標題: 多協定標籤交換路徑恢復
MPLS PATH RESTORATION
作者: 謝界民
Hsieh, Chieh-Min
關鍵字: 多協定標籤交換;MPLS
出版社: 電機工程學系
摘要: 
今日之通訊網路及服務都趨向集中在IP 網路,多協定標籤交換 Multiprotocol Label Switching (MPLS) 融合了IP 路由的彈性及 Layer2 交換的效能已廣為大家所接受,預估未來 MPLS 是下一代 IP 骨幹網路的主流,MPLS 技術已證明它的價值做為運送新的服務並允許在同時間轉移舊網路到新網路。MPLS 的故障恢復機制的技術是對一失敗或擁塞的 LSP(Label Switched Path) 進行重繞流量。在 Haskin’s 的提議中提供一個快速重繞保護機制為一個在MPLS網路中失敗的主要/被保護的LSP路徑設定一個替代LSP路徑去處理快速重繞流量。此提議提供一個近乎50毫秒的恢復時間,但卻有兩個問題: 其一是當ingress LSR 偵測到流量在反向路徑它會直接交換到替代LSP,當封包從ingress LSR 到最後失敗的那一點的LSR 運送流量回到反向路徑迴路如此封包將會有兩路延遲問題,另一問題是當封包要從替代 LSP 路徑要回到原來主要LSP 路徑時會與反向封包混合產生封包失序問題。 在Makam’s 的提議中使用一個故障偵測通知機制去運送有關一個故障的節點訊息以便讓ingress LSR 直接導向流量到一預設的替代LSP路徑。此方法提供比Haskin’s 的機制更好的資源使用率,但流量傳送卻在偵測故障通知訊息到達 ingress LSR 之期間會被通告的LSR 丟棄。本篇論文是改良自Hskin’s 的保護交換機制及結合Makam’s 的機制提出更有效的重繞機制改善封包失序的問題,支援多協定標籤MPLS路徑恢復的失敗類別比如鏈路失敗或節點失敗, 經過模擬與現有機制做一比較。
關鍵詞:MPLS,多協定標籤路徑恢復。

Today's communication networks and services are migrating to a converged paradigm centered on IP (Internet Protocol). MPLS (Multiprotocol Label Switching) has recently been accepted as a new approach for integrating IP layer 3 routing with layer 2 switching technology. It is estimated that in the near future, MPLS will be chosen as bearer of IP large backbone networks. MPLS technology has proven its value for delivering new services while at the same time allowing migration from old to new networks. The path recovery in MPLS is the technique to reroute traffic around a failure or congestion in a LSP(Label Switched Path). In Haskin's proposal the authors present a fast rerouting protection mechanism for setting up an alternative LSP to handle fast rerouting of traffic upon a failure in primary/protected LSP in MPLS network. This proposal provides a quick restoration time nearly to the 50 milliseconds. But this proposal with the two problem: One is while the ingress LSR detects traffic in the reverse direction it switches the traffic directly to the alternative LSP. These packets will with a two-way delay while traversing the backwards loop from the ingress LSR to the last LSR at the point of failure. Another problem of this scheme is that as packets arriving from the reverse direction are mixed with incoming packets, this results in packet disordering through the alternative LSP during the restoration period. In Makam's proposal the authors use a fault notification mechanism to convey information about the occurrence of a fault to a responsible node in order to let ingress LSR redirects the traffic over a pre-established alternative LSP. This method provides better resource utilization the Haskin's scheme. But the the traffic that is in transit during the interval of time between the detection of the fault detected and the time the fault notification signal reaches the ingress LSR will be droped by the alert LSR. My proposal improve Haskin's protection switching and combined with Makam's scheme proposes an efficient rerouting scheme can decrease the packets reordered prolem , support MPLS path restoration failure types such as link failures, node failures. Through simulation, the performance of the proposed scheme is measured and compared with the existing schemes.
Keyword: MPLS, MPLS path restoration.
URI: http://hdl.handle.net/11455/7649
Appears in Collections:電機工程學系所

Show full item record
 

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.