Green Communications and Networking. First International Conference, GreeNets 2011, Colmar, France, October 5-7, 2011, Revised Selected Papers

Research Article

Gradient Optimisation for Network Power Consumption

Download57 downloads
  • @INPROCEEDINGS{10.1007/978-3-642-33368-2_11,
        author={Erol Gelenbe and Christina Morfopoulou},
        title={Gradient Optimisation for Network Power Consumption},
        proceedings={Green Communications and Networking. First International Conference, GreeNets 2011, Colmar, France, October 5-7, 2011, Revised Selected Papers},
        proceedings_a={GREENETS},
        year={2012},
        month={11},
        keywords={},
        doi={10.1007/978-3-642-33368-2_11}
    }
    
  • Erol Gelenbe
    Christina Morfopoulou
    Year: 2012
    Gradient Optimisation for Network Power Consumption
    GREENETS
    Springer
    DOI: 10.1007/978-3-642-33368-2_11
Erol Gelenbe1,*, Christina Morfopoulou1,*
  • 1: Imperial College
*Contact email: e.gelenbe@imperial.ac.uk, c.morfopoulou@imperial.ac.uk

Abstract

The purpose of this paper is to examine how a gradient-based algorithm that minimises a cost function that includes both quality of service (QoS) and power minimisation in wired networks can be used to improve energy savings with respect to shortest-path routing, as well as against a “smart” autonomic algorithm called EARP which uses adaptive reinforcement learning. Comparisons are conducted based on the same test-bed and identical network traffic. We assume that due to the need for network reliability and resilience we are not allowed to turn off routers and link drivers. We also assume that for QoS reasons (notably with regard to jitter and to avoid packet desequencing) we are not allowed to split traffic from the same flow into different paths. Under these assumptions and for the considered traffic, we observe that power consumed with the gradient-optimiser is a few percent to 10% smaller than that consumed using shortest-path routing or EARP. Since the magnitude of the savings is small, this suggests that further power savings may only be obtained if under-utilised equipment can be dynamically put to sleep or turned off.