Concepts and Research into Green Networking

Green Networking

Abstract:

Saving of excessive energy consumption is becoming a key concern in networking, because of theprobable economical benefits. These concerns, usually argued to as “green networking”, relate to inserting energy-awareness in the strategy, in the devicesand in the protocols of networks. In this work, I first formulate a more precise definition of the “green” attribute. I further more classify a few standards that are key enablers of energy-aware networking research. I then overview the up-to-date state of the art and offer a catalog of the relevant work, with a superior focus on networking.

Introduction:

Recent studies shows that Information Technology and communication advances are responsible for significant amount of world electric power consumptions which ranges from 2% to 10%, that is one of the contributing factor for global warming, via gases release from greenhouse and from the growth of demand of internet applications and services. Therefore, for these reasons energy efficient and sustainable networking often called “Green Networking”, has become a hot issue in the last few years.

Definition: Green Networking is the practice of selecting energy-efficient networking technologies and products, and minimizing resources use whenever possible. [1]

All facets of Information Technology and Communication are under supervision, from energy-saving design of all networking devices, to strategies which reflect the entire network’s energy depletion in the planning, design, implementation and management points, to new approach for long-run sustainability of the networking which covers reformed attitudes of users’ as well as smart energy mowing techniques.This special concern on Green Networking intentions at providing revolutionary influences to the research and development of energy-efficient networking solutions and approaches for network sustainability.

Motivation for research and review:

There are many complexities in network, unnecessary traffic congestion, and more power wastages due to unnecessary idling of node at the current situation of networking.Therefore, I have used one research paper published on the subject and examining the viewpoints to the matter.To the paper, I have added an analysis section where I reflect on the paper value, and discuss important points listed in the paper on the subject.

Green Networking With Packet Processing Engines: Modeling and Optimization

Published on:14^th February, 2013. IEEE Computer Society

Authors: Raffaele Bolla, RobertoBruschi, Alessandro Carrega, and Franco Davoli

With the goal of monitoring power consumption in metro/transport and main networks, the paper reflects energy-aware devices capable to shrink their energy chucks by adjusting their performance. In specific, the paper focuses on state-of-the-art packet processing engines, which normally characterize the most energy-consuming apparatuses of network devices, and that are often collected of a number of parallel pipelines to “divide and conquer” the received traffic load. The paper talk about goal to control both the power structure of pipelines and the way to issue traffic flows among them. The authors proposed an analytical model to precisely represent the impact of green network technologies (i.e., low power idle and adaptive rate) on network-aware and energy-aware performance indexes. The model has been confirmed with experimental consequences, accomplished by using energy-aware software routers loaded by real-world traffic traces. The attained outcomes determine how the projected model can successfully epitomize energy-aware and network-aware presentation indexes. The method goals at dynamically adjusting the energy-aware device structure to lessen energy consumption whereas handling with received traffic signals and gathering network performance limitations. In order to genuinely comprehend the impact of such policy, a number of experiments have been executed by using experimental data from software router designs and real-world traffic traces.

Analysis:

In this paper, the authors considered energy-aware network devices (e.g., routers, switches, etc.) able to trade their energy consumption for packet forwarding performance by means of both low power idle and adaptive rate schemes. The proposed analytical model is able to capture the impact of power management capabilities on network performance metrics. The analytical framework considers stochastic incoming traffic at the packet level with Long Range Dependency (LRD) properties. On the basis of the analytical model, authors have chosen the parameters characterizing the joint usage of Adaptive Rate(AR) and Low Power Idle(LPI) energy-aware capabilities by optimizing the desired tradeoff between energy consumption and Quality of Service(QoS) while at the same time enforcing the satisfaction of given upper bounds on both. Since the performance and cost indicators used in the optimization depend on incoming traffic volumes and statistical features (notably, burst inter-arrival time and average burst length), researchers repeat the optimization periodically under updated estimations of these quantities. The modeling and control framework has been validated experimentally by using a Linux-based open software router with AR and LPI primitives under traffic generated by real-world traces; the results demonstrate how the proposed model can effectively represent energy-aware and network-aware performance indexes. Therefore proposed model, is efficient and addressing green networking maintaining the Quality of Service (QoS) in the network.

Green Strategies

Traditionally, networking systems are designed and dimensioned according to principles that are inherently in opposition with green networking objectives: namely, over-provisioning and redundancy. On the other hand, due to the lack of Quality of Service (QoS) provision from the Internet architecture, over-provisioning is a common practice: networks are dimensioned to sustain peak hour traffic, with extra capacity to allow for unexpected happenings. As a result, through low traffic periods, over-provisioned networks are also over-energy-consuming. Furthermore, on behalf of resiliency and fault-tolerance, networks are also deliberate in a redundant manner. Devices are added to the structure with the sole purpose of taking over the duty when another device fails, which further adds to the overall energy ingesting. These objectives, drastically divergent to the environmental ones, make green networking an interesting, and technically challenging, research arena. A major change is indeed needed in networking research and development to introduce energy-awareness in the network design, deprived of compromising either the Quality of Service (QoS) or the network consistency. This section illustrates a few key paradigms that the network infrastructure can exploit to reach the green objectives formalized above. We individuate three classes of solution, namely resource consolidation, virtualization and selective connectedness [2]. These three categories represent three research directions, which may find further detailed applications in device and protocol design.

Resource consolidation regroups all the dimensioning strategies to reduce the global consumption due to devices underutilized at a given time. Given that the traffic level in a given network approximately follows a well-known daily and weekly behavior [3], there is an opportunity to “adapt” the level of active over-provisioning to the current network conditions. In other words, the required level of performance will still be guaranteed, but using an amount of resources that is dimensioned for current network traffic demand rather than for the peak demand. This can, for example, be achieved by shutting down some lightly loaded routers and rerouting the traffic on a smaller number of active network equipment. Resource consolidation is already a popular approach in other fields, in particular data centers and CPU.

Virtualization regroups a set of mechanisms allowing more than one service to operate on the same section of hardware, hence refining the hardware operation. It results in a lowered energy consumption, as long as a single machine under high load consumes less than several lightly loaded ones, which is generally the case. Virtualization can be applied to multiple kinds of resources, comprising network links, storage hardware, software resources, etc. A typical example of virtualization consists in sharing servers in data centers, thus reducing hardware costs, improving energy management and reducing energy and cooling costs, ultimately reducing data center carbon footprint. In the current context, virtualization has already been deployed with success: e.g., the US Postal Service has virtualized 791 of its 895 physical servers [4]. As virtualization is a more mature research field, we refer the interested reader to [5] for a detailed survey of virtualization techniques from a computer architecture perspective, and to [6] for a networking perspective. At the same time, it should be noted that a virtualization solution designed explicitly to reduce network energy consumption has yet to appear.

Applying the same base concept, selective connectedness of devices, as outlined in [7], [8], consists in distributed mechanisms allowing single pieces of equipment to go idle for some time, as clearly as probable for the rest of the networked devices. If the consolidation principle relates to resources that are shared within the network infrastructure, selective connectedness allows instead to turn off unused resources at the edge of the network. For instance, edge nodes can go idle in order to avoid supporting network connectivity tasks (e.g., periodically sending heartbeats, receiving unnecessary broadcast traffic, etc.). These tasks may have to be taken over by other nodes, such as proxies, momentarily faking identity of sluggish devices, so that no essentialmodification is required in network protocols

Conclusion:

This article surveyed the efforts that the research community has been spending in the attempt to reduce the energy waste in fixed networks, which are usually denoted as “green networking”. I presented the importance of the issue, its definition and mainstream paradigms, and proposed a taxonomy of the relevant related work. Examining the state of the art, we observe that a few techniques are emerging, which can be roughly categorized as (i) resource consolidation, (ii) virtualization (iii) selective connectedness. It also emerges from my analysis that despite the relative youth of the green networking field, research in some of the above areas is already mature, with advanced standardization efforts and prototyping results.

Finally, as the ultimate goal of networking is to provide services to end-users, the quality of such services and of the user experience is a topic that spans over all the previous branches. Indeed, while energy efficiency is becoming a primary issue, it shall never be neglected that the energy gain must not come at the price of a network performance loss. This delicate tradeoff arises from opposite principles: indeed, while networked systems have traditionally be designed and dimensioned according to principles such as overprovisioning and redundancy, green networking approaches praise opposite practices such as resource-consolidation and selective-connectedness. The challenge lays in this case in applying the latter principles in a way that is as transparent as possible to the user – in other words, avoiding that resource consolidation translates into congestion, or that selective connectedness translates into unreachability. While the first wave of green studies focused more on the achievable energy gain, we believe that a systematic evaluation of networking performance from the user-perspective should be undertaken as well. Indeed, in all branches interesting questions remain, which deserve precise quantitative answers: Finally, we believe that while, for the time being, techniques of different branches have been studied in isolation, future research should address the combined impact of different techniques as well. Indeed, even though each of the above techniques alone do not constitute serious threats for the QoS perceived by the end-user, however it is not guaranteed that the joint use of several technique will not raise unexpected behaviour. Due to the current rise in green networking research and attention, it cannot be excluded that, in a near future, users will run Energy Aware Applications, in a home equipped with a green set-top-box implementing Interface Proxying functionalities, and will access the Internet through an Internet Service Provider implementing Energy Aware Routing in devices interconnected by Adaptive Link Rate lines – which opens a number of interesting questions that are so far all unexplored.

References:

[1] “Discussion – Definition of green networking,”http://searchnetworking.techtarget.com/definition/green-networking

[2] “International Journal of Advanced Research in Computer Science and Software Engineering,”http://ijarcsse.com/docs/papers/Volume_4/9_September2014/V4I9-0353.pdf

[3] A. Qureshi, R. Weber, H. Balakrishnan, J. Guttag, and B. Maggs, “Cutting the Electric Bill for Internet-Scale Systems,” in Proceedings of the ACM Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications (SIGCOMM 2009), (Barcelona, Spain), Aug. 2009.

[4] U.S. Environmental Protection Agency – Energy Star Program, “Report to Congress on Server and Data Center Energy Efficiency Public Law 109-431,” Tech. Rep. , U.S. Environmental Protection Agency, Aug. 2007.

[5] S. Nanda and T.-C. Chiueh, “A Survey on Virtualization Technologies,” Tech. Rep. TR179, Department of Computer Science, SUNY at Stony Brook, 2005.

[6] N. M. Kabir Chowdhury and R. Boutaba, “A Survey of Network Virtualization,” Tech. Rep. CS-2008-25, University of Waterloo, Oct.2008.

[7] K. Christensen, C. Gunaratne, B. Nordman, and A. D. George, “The Next Frontier for Communications Networks: Power Management,”Computer Communications, vol. 27, pp. 1758–1770, Dec. 2004.

[8] M. Allman, K. Christensen, B. Nordman, and V. Paxson, “Enabling an Energy-Efficient Future Internet Through Selectively Connected End Systems,” in Proceedings of the Sixth ACM Workshop on Hot Topics in Networks (HotNets-VI), (Atlanta, Georgia, USA), Nov. 2007.