Energy-efficient bandwidth reservation for bulk data transfers in dedicated wired networks

Abstract

International audienceThe ever increasing number of Internet connected end-hosts call for high performance end-to-end networks leading to an increase in the energy consumed by the networks. Our work deals with the energy consumption issue in dedicated network with bandwidth provisionning and in-advance reservations of network equipments and bandwidth for Bulk Data transfers. First, we propose an end-to-end energy cost model of such networks which described the energy consumed by a transfer for all the crossed equipments. This model is then used to develop a new energy-aware framework adapted to Bulk Data Transfers over dedicated networks. This framework enables switching off unused network portions during certain periods of time to save energy. This framework is also endowed with prediction algorithms to avoid useless switching off and with adaptive scheduling management to optimize the energy used by the transfers. 1 Introductio

A survey on techniques for improving the energy efficiency of large-scale distributed systems

“…However, these solutions require major changes such as dynamic routing protocols that can handle shutdown nodes and links, a centralized management to coordinate switch off [Gelenbe and Silvestri 2009] and higher-level cooperation between ISPs and Content Providers (CP) [Chiaraviglio and Matta 2010]. Bandwidth provisioning frameworks can be adapted to · 27 save energy [Orgerie and Lefèvre 2011]. Network management architectures have also been proposed to enforce business decisions aiming to achieve energy efficiency by exploring the trade-off between energy savings and network performance [Costa et al 2012].…”

Dynamic Adaptive Management of Buffer for QoS Guarantees and Energy Reduction

“…This module provides the instantaneous energy consumption of each equipment taking into account the traffic and the type of employed equipment (router, switch, network card, etc.) [9]. This module is based on the ALR model (Adaptive Link Rate) which adapts transmission rates depending on link use-rate.…”

“…The adaptive law controls the traffic that crosses a node, by using the length-queue where packets are temporarily stored.This proposed law guarantees unsaturated output traffic to avoid packet loss, which ensures, simultaneously, a fast convergence to a reference level. This reference level in the buffer is given by the energy thresholds proposed by the ALR model, which ensures energy reduction [9].…”

Buffer dynamic management for energy-aware network

“…This module provides the instantaneous energy consumption of each equipment taking into account the traffic and the type of employed equipment (router, switch, network card, etc.) [24]. This module is based on the ALR model (Adaptive Link Rate) which adapts transmission rates depending on link use-rate.…”

“…trade off between energy saving and quality of service parameters such as transmission delay information). Likewise, energy consumption measures are given by ECOFEN [24] (implemented module in NS-2).…”