軟體定義網路群播演算法效能比較

Performance Comparisons of Multicast Algorithms for Software Defined Networking

馬哈德(Mahardeka Tri Ananta)

指導教授 ： 江振瑞

國立中央大學/資訊電機學院/資訊工程學系/碩士(2014年)

摘要

軟體定義網路(Software Defined Networking, SDN)的概念係網路設備的控制面(control plane)與資料面(data plane)分離，控制器(controller)以集中控管的方式設定各交換器的流表(flow tables)，而由交換器透過流表負責轉傳資料封包。SDN使用者可以透過撰寫控制器上的應用程式對整個網路進行監督與管理。於SDN中，一個典型的應用為群播(multicast)，而耗費網路中大量頻寬的多媒體串流資料(如影片、音訊)則是群播應用的主要資料來源。我們於本論文中設計並改善群播演算法(或應用程式)，使其運行於控制器之上並產生群播樹，利於發行者(publisher)將資料封包發送給訂閱者(subscriber)，使得該群播樹中所有節點與所有主機接收到每個封包至多一次，以降低網路頻寬的消耗。我們使用Pyretic於SDN中實作三個群播演算法，它們分別基於廣度優先搜尋樹(breadth-first search tree, BFST)、Dijkstra 最短路徑樹(Dijkstra’s shortest path tree, DSPT)和延伸Dijkstra 最短路徑樹(extended Dijkstra’s shortest path tree, EDSPT) ，並使用Mininet模擬器在Abilene拓樸(Abilene Topology)的環境下比較上述演算法的效能。模擬比較結果顯示，使用EDSPT 的群播演算法具有最高的吞吐量(throughput)、最低的抖動率(jitter)以及封包遺失(packet loss)。

關鍵字: 軟體定義網路 (SDN)、 群播、廣度優先搜尋、Dijkstra最短路徑演算法

關鍵字

軟體定義網路 ； 群播 ； 廣度優先搜尋 ； 最短路徑演算法

並列摘要

Software Defined Networking (SDN) is a concept to decouple the control plane and data plane of network devices. In SDN, a logically centralized controller configures the forwarding tables (also called flow tables) of switches, which are responsible for forwarding the packets of communication flows. In this way, SDN users can composite application programs run on top of the controller to monitor and manage the whole network. One of typical SDN applications is multicast. Multimedia data (e.g., video and audio data) have long been a major source in communication networks that consumes a lot of bandwidth. This thesis aims to design and implement multicast algorithms (or applications) to run on top of the controller to generate a multicast tree for a data publisher to deliver data packets to all subscribers so that every node and every host on the multicast tree will receive every packet once and at most once for reducing bandwidth consumption. We use Pyretic to implement three multicast algorithms over a SDN network, based on constructing the breadth-first search tree (BFST), the Dijkstra’s shortest path tree (DSPT), and the extended Dijkstra’s shortest path tree (EDSPT), respectively, and compare their performance under the Abilene Topology with the Mininet network simulator. As shown by the comparisons, the algorithm using EDSPT achieves the best performance in terms of throughput, jitter, and packet loss.

Keywords: Software Defined Networking (SDN); Multicast; Breadth-First Search; Dijkstra’s Shortest Path Algorithm

並列關鍵字

Software Defined Networking ； Multicast ； Breadth-First Search ； Dijkstra’s Shortest Path Algorithm

參考文獻

[*SPT] Shortest Path Tree . http://www.csie.ntu.edu.tw/~kmchao/tree07spr/spt.pdf , last accessed on June 2014

[AKL13] Sugam Agarwal, Murali Kodialam, T. V. Lakshman, "Traffic Engineering in Software Defined Networks," Proceedings IEEE INFOCOM, Bell Labs Alcatel-Lucent Holmdel, 2013.

[DTS13] Saurav Das, Et al., “Handbook of Fiber Optic Data Communication a Practical Guide to Optical Networking Chapter 17, 4th edition”.

[JHJS2014] Jehn-Ruey Jiang, Hsin-Wen Huang, Ji-Hau Liao, and Szu-Yuan Chen, "Extending Dijkstra’s Shortest Path Algorithm for Software Defined Networking," Technical Report, National Central University, 2014.

[KF13] Hyojoon, Kim Feamster, N, “Improving Network Management With Software Defined Networking,” Communications Magazine, IEEE, 2013.