A novel utilisation-aware energy consumption model for content distribution networks

Abstract

Rapid increase in internet worldwide users is leading a trend towards grand geographical distributed systems. Content distribution networks (CDNs) are one of the popular large-scale distributed systems managing around 40% of internet traffic. In order to reduce services cost and to minimise internet contribution in we footprint, it is very important to manage resources efficiently and to control energy consumption in CDNs. Although different studies propose solutions to reduce energy consumption, there is less concentration on analysing energy consumption and resource utilisation. In this paper, we have evaluated resource utilisation and ultimately energy consumption under different scenarios. We have proposed utilisation-aware energy consumption model. We have considered one of the most popular request redirection policy, that is, load balance. Quality of experience (QoE) metrics such as overall delay in client requests completion and dropped requests are also evaluated. Extensive simulation is performed by varying CDN infrastructure size, client requests traffic volume and intensity of end-user requests.