AttentiveSDN: EndHost Awareness-Based Power-Optimized Software-Defined Networks

AttentiveSDN: EndHost Awareness-Based Power-Optimized Software-Defined Networks

Mahmoud Al Ahmad, Suchismita Rout, Sudhansu Shekhar Patra, Bibhudatta Sahoo, Harishchandra Dubey, Rabindra Kumar Barik
Copyright: © 2022 |Pages: 22
DOI: 10.4018/JITR.299381
Article PDF Download
Open access articles are freely available for download

Abstract

Internet users are increasing day by day due to its support for many applications and creation of innovative services. Along with this, energy consumption is also becoming an important concern in networking. Several researchers have investigated energy saving schemes for networks. Software Defined Networking (SDN) is an excellent choice which improves network functionalities with flexible management aided by centralized control. Recent studies designed efficient algorithms for advancing SDN with overall energy savings. Shutting down idle links and switches are one among numerous such solutions available for SDN. In this paper, we proposed a novel algorithm named AttentiveSDN for reducing the energy consumption in SDNs. Here the controller collects the traffic and the link status from the switches involved in network operation and takes the decision to put which idle links and switches into sleep state . We evaluated the performance of the AttentiveSDN algorithm using Mininet. The result has shown that our proposed approach saves more power than existing solutions.
Article Preview
Top

1. Introduction

A lot of researches have been conducted in order to save battery energy in a wireless network as its battery power is limited in nature. In order to maximize network lifetime energy expenditure of the wireless network should be minimized. Now a day’s researchers put their attention in wired network also to minimize the energy utilization of the whole network (Molina et al., 2018). Software Defined Networking (SDN) is an important name in the field wired network. The popularity of SDN network is due to its separation of control and data plane of the network. It provides the application with an abstracted distributed state in a centralized manner. Figure 1 (a) and Figure 1(b) are illustrated the traditional network and general architecture of SDN respectively. This unique characteristic of controlling the network centrally has attracted many companies such as Amazon, Google, Microsoft etc. to deploy SDN network as its backbone network.

The three-layer structure of SDN architecture has shown in Figure (2). With the rise in popularity of this network energy consumption of this network is also increasing at an alarming rate (Jia et al.,2018). The energy consumption puts a serious limitation on SDN. The internet energy consumption is almost 10% of the total energy consumption of the world (Assefa et al.,2015) There are many reasons to focus on energy saving characteristics of wired networks like SDN. The former one is due to environmental alertness. This is a global alarming and pressure from both political and social points. Another reason is due to the tremendous growth of internet traffic, which is increasing rapidly every year. To increase in internet traffic there is also an incremental growth of network energy consumption. Therefore, this terrific amount of energy consumption is not only an issue as per the environment or society is considered, but also it is frightening due to rapid growth and popularity on the internet, due to energy requirement for running of the network equipment and heat dissipation in data centers. The amount of energy consumption in 2010 is estimated as 21.4 TWH in European Telecom by Global E-Sustainability Initiative. At the same time, it also forecasts that the figure of energy consumption in 2020 as 35.8 TWH if no initiatives for power savings is taken into consideration (Webb et al.,2008).

Unlike wireless network, in wired networks, there is no burden on the networking components, to find out the route(Sahoo et al.,2016; Bhushan & Sahoo, 2018 ; Bhushan & Sahoo, 2019 ; Bhushan & Sahoo, 2020). Each router and switch in wired network are connected with the links. In order to construct a fully robust network link bandwidth and redundancies are adopted. Which result in more power consumption of the link. It is observed in wired link that whether the link is idle or fully utilized the amount of energy spent in the link is same. Energy consumption in the switch is not relative to the utilization of its links, which implies that energy consumption in the switch has no relation with incoming traffic in the network. In the network it is found that almost all the switches are active at all time, but only 40% of the link capacities is used, whereas by turning off the unutilized switch and its corresponding links can save most of the network power. A switch consumes most of its power when it is on, whereas in the transition period from idle state to fully utilize will consume only 8% of power (Bista et al.,2014). Therefore, to save the maximum amount of energy in the network proper power measures must be taken in order to identify the unutilized links and switches and making them turn off. Energy savings can be achieved in a precise manner by controlling the flow table of the forwarding path from the controller. It is also observed that in wired network energy consumption is more with high link speed. Energy consumption of 1 Gbps Ethernet link uses 4W more than 100 Mbps link.

Complete Article List

Search this Journal:
Reset
Volume 16: 1 Issue (2024): Forthcoming, Available for Pre-Order
Volume 15: 6 Issues (2022): 1 Released, 5 Forthcoming
Volume 14: 4 Issues (2021)
Volume 13: 4 Issues (2020)
Volume 12: 4 Issues (2019)
Volume 11: 4 Issues (2018)
Volume 10: 4 Issues (2017)
Volume 9: 4 Issues (2016)
Volume 8: 4 Issues (2015)
Volume 7: 4 Issues (2014)
Volume 6: 4 Issues (2013)
Volume 5: 4 Issues (2012)
Volume 4: 4 Issues (2011)
Volume 3: 4 Issues (2010)
Volume 2: 4 Issues (2009)
Volume 1: 4 Issues (2008)
View Complete Journal Contents Listing