A Low-Cost Methodology for Profiling the Energy Consumption of Network Equipment

Rate adaptation technologies aim at establishing a linear relationship between power consumption and traffic load in packet networks. They rely on energy profiles of network equipment, which map system configurations and traffic loads onto power consumption levels, for selection of network resources to place into low-power states and identification of new system designs with high energy-saving yields. We define a new methodology for profiling the energy consumption of network systems that reconciles modeling accuracy with cost containment and rapidity in the preparation and execution of the power measurements. We apply the methodology to five systems from multiple vendors, two of which from an off-site location. Our results show that some energy savings can derive from straightforward signaling extensions and system-management software upgrades that support demand-timescale rate adaptation techniques, but more substantial savings can only be attained with a new generation of hardware platforms where packet-timescale rate adaptation is pervasively deployed.