Analytical Model for OpenFlow-Based Software-Defined Network

被引：5

作者：

Sarkar, Chiranjit ^{[1
]}

Setua, S. K. ^{[1
]}

机构：

[1] Univ Calcutta, Dept Comp Sci & Engn, Kolkata, India

来源：

PROGRESS IN COMPUTING, ANALYTICS AND NETWORKING, ICCAN 2017 | 2018年 / 710卷

关键词：

OpenFlow; SDN; M/M/1 queueing model;

D O I：

10.1007/978-981-10-7871-2_56

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Software-defined network (SDN) is an emerging network architecture that can be adapted as a building block for cloud-enabled data centers. SDN decouples the traditional network in Data Plane (collection of forwarding devices or OpenFlow-enabled switches) and Control Plane (collection of one or more controllers), which communicates with each other using OpenFlow protocol. Proper installation of SDN necessitates the performance test of OpenFlow-enabled switches and controllers. Efficiency, reliability, and scalability are few parameters that should be considered to calculate the efficiency of an OpenFlow-based SDN. An analytical model of the OpenFlow-based network is suggested in this paper. Our proposed model is based on queueing theory and exponential models. The average packet arrival rate of a controller and maximum waiting time for an OpenFlow-enabled switch to get a PACKET_OUT message are taken as parameters to define our approach.

引用

页码：583 / 592

页数：10