The main motivation of this work is to investigate techniques to reduce the power consumption inside a network element. It is enough to consider the high energy demand associated to the telecommunication networks field. As practical consequence the power consumption has become a relevant parameter and it represents a critical constraint for the network designers looking both the whole network infrastructure and the network elements like switches, routers and servers. The PhD has been focused mainly on two research areas of interest, the first one was the power consumption inside the switching fabric of an high speed router. The target was to analyze the effect of the dynamic power inside a switching fabric, to evaluate a set of optimization strategies in order to minimize the power consumption and to achieve the best trade-off between power, high performances and packet delays; the crossbar was used as reference switching architecture for this study. Looking at the consumption side, generally speaking, it is possible to define two families of switching fabrics: 1)Bit-rate independent switching fabric, in which the consumption does not depend on the number of transported bits; this family is typical of optical switching fabrics 2)Bit-rate dependent switching fabric, where the total consumption is proportional to the data transmission bit-rate, this family is typical of electronic switching fabrics The second research activity was carried at the Alcatel-Lucent Bell Laboratories, based in New Jersey (USA) and over a period of 9 months. The study of the power consumption across several network elements that are commercially available for the "corporate" market. We started from a set of collected larger number of power measurements over these network elements and thanks to them we were able to develop a linear mathematical model to describe the power consumption of a generic network element.

Energy aware control algorithms for computer networks / Ricca, Marco. - (2012). [10.6092/polito/porto/2497193]

### Energy aware control algorithms for computer networks

#### Abstract

The main motivation of this work is to investigate techniques to reduce the power consumption inside a network element. It is enough to consider the high energy demand associated to the telecommunication networks field. As practical consequence the power consumption has become a relevant parameter and it represents a critical constraint for the network designers looking both the whole network infrastructure and the network elements like switches, routers and servers. The PhD has been focused mainly on two research areas of interest, the first one was the power consumption inside the switching fabric of an high speed router. The target was to analyze the effect of the dynamic power inside a switching fabric, to evaluate a set of optimization strategies in order to minimize the power consumption and to achieve the best trade-off between power, high performances and packet delays; the crossbar was used as reference switching architecture for this study. Looking at the consumption side, generally speaking, it is possible to define two families of switching fabrics: 1)Bit-rate independent switching fabric, in which the consumption does not depend on the number of transported bits; this family is typical of optical switching fabrics 2)Bit-rate dependent switching fabric, where the total consumption is proportional to the data transmission bit-rate, this family is typical of electronic switching fabrics The second research activity was carried at the Alcatel-Lucent Bell Laboratories, based in New Jersey (USA) and over a period of 9 months. The study of the power consumption across several network elements that are commercially available for the "corporate" market. We started from a set of collected larger number of power measurements over these network elements and thanks to them we were able to develop a linear mathematical model to describe the power consumption of a generic network element.
##### Scheda breve Scheda completa Scheda completa (DC)
2012
File in questo prodotto:
File
mrc-tesi.pdf

Open Access dal 01/01/2023

Tipologia: Tesi di dottorato
Licenza: Creative commons
Dimensione 1.06 MB
Utilizza questo identificativo per citare o creare un link a questo documento: `https://hdl.handle.net/11583/2497193`