ERCIM news 135
ERCIM news 135
ERCIM news 134
ERCIM news 134
ERCIM news 133
ERCIM news 133
ERCIM news 132
ERCIM news 132
ERCIM news 131
ERCIM news 131
ERCIM news 130
ERCIM news 130
Back Issues Online
Back Issues Online

by David Hausheer, University of Zurich, Switzerland

The EcoNets project aims to develop new mechanisms for an economic and energy-efficient bandwidth allocation in future networks. Cooperation and resource sharing combined with new allocation schemes will provide appropriate incentives to provide bandwidth only where and when it is needed.

There has recently been tremendous technological progress in the area of Internet-based communications, with router hardware now able to process data at a rate of several terabits per second. At the same time, the proliferation of wireless technology has enabled users to be ‘connected’ anytime and anywhere in the world. However, it is important to consider all the costs that arise from such new technology. Besides acquisition and maintenance costs, energy consumption is becoming increasingly significant. Rising energy costs and the need to reduce carbon emissions have led to an increased awareness of the need to improve energy efficiency wherever possible. For computing equipment such as monitors and desktop computers, standards like EnergyStar exist to ensure that energy resources are used efficiently. However, little attention has so far been paid to energy consumption in networks, and comparable standards for networking equipment do not yet exist. This is despite the fact that energy consumption of network devices is a substantial and growing cost factor due to higher-capacity equipment that consumes more energy - up to 50% of the total cost of ownership according to some estimates.

Fortunately, there is potential to improve energy efficiency in networks, as despite their energy consumption being high, their utilization is generally below 5%. Access points of wireless networks are idle much of the time, and consume energy whether or not they are in use. This energy consumption could be reduced by cooperation between different wireless network providers.

Recent advances in the area of network virtualization enable the provisioning of bandwidth ‘on demand’. Besides providing numerous other benefits relating to security, flexibility and reliability, this technology enables the transparent sharing of physical network equipment between different customers of the same network provider. An important aspect of network virtualization that has received little attention so far is its potential to significantly reduce energy costs, as it allows customers to gradually adapt bandwidth capacity to current demand. However, suitable business models for such on-demand bandwidth services have not yet evolved, and without appropriate settlement schemes providers have few incentives to cooperate.

EcoNets Objectives
EcoNets is a postdoctoral project supported by a fellowship from the Swiss National Science Foundation (SNSF) in collaboration with the University of California at Berkeley. The project, which began in late 2009, aims to develop new incentive schemes and allocation mechanisms for economic and energy-efficient bandwidth provisioning in fixed and wireless network infrastructures of the future. The following key economic and technical research issues have been identified and will be addressed in EcoNets:

  • Energy-efficient bandwidth provisioning: A key part of bandwidth provisioning costs is the energy cost, making minimum consumption of energy resources essential. Even small energy savings can have a tremendous impact if summed over a huge number of networking devices. This objective is in line with EU plans to reduce primary energy use by 20% before 2020. However, this ambitious goal can only be reached if energy efficiency is improved on all levels.
  • Bandwidth on demand: New bandwidth allocation schemes and business models need to be developed to support the on-demand provision of bandwidth services to customers at the right location, at the right time and in the right quantity. However, the effect of such bandwidth-on-demand allocation schemes on higher-level routing and transport protocols needs to be investigated.
  • Economically efficient incentives: The design of resource allocation policies and incentives for cooperation in this context is a difficult problem. Incentive mechanisms should lead, ideally, to an economically efficient allocation and use of bandwidth services. Auctions are a standard way to achieve such objectives, but the distributed environment and the different types of resource involved pose significant challenges to their implementation.
  • Scalable and robust bandwidth allocation: Any bandwidth provisioning scheme needs to be able to scale to a very large amount of bandwidth offered by many providers to many customers, and over a very large number of network devices. The support of bandwidth allocation in a fully decentralized manner, such as based on peer-to-peer (P2P) concepts (eg PeerMart), shows advantages in terms of robustness and scalability for large systems.

EcoNets will design allocation mechanisms to allow customers to reserve bandwidth capacity on demand and in advance. Reserved capacity that remains unused may later be offered to other customers. Thus, customers can themselves become providers of bandwidth services. The key architectural idea is to organize the different network entities in an overlay network in which they become peers (cf. Figure 1). This peer awareness will allow an optimal allocation of bandwidth to be reached. Customers will benefit from these mechanisms as their bandwidth needs can be met in a more scalable and cost-effective manner. Finally, by following a generic bandwidth allocation approach, EcoNets will be applicable to both fixed and wireless network scenarios.

Links:
EcoNets Project: http://www.csg.uzh.ch/research/econets
PeerMart Homepage: http://www.peermart.net/
Dagstuhl Seminar on Bandwidth on Demand: http://www.dagstuhl.de/09072

Please contact:
David Hausheer
University of Zurich, IFI, Switzerland
Tel: +41 44 635 43 72
E-mail: This email address is being protected from spambots. You need JavaScript enabled to view it.

Next issue: January 2024
Special theme:
Large Language Models
Call for the next issue
Get the latest issue to your desktop
RSS Feed