A control theoretic approach for energy-efficient management of Online Social Network services

Abstract

As a consequence of the massive utilization of Online Social Networks (OSNs) they must exploit a huge amount of hardware and network appliances. Therefore, optimizing their energy consumption could lead to major benefits both from the environmental and economical viewpoints. In this perspective, this paper defines a general OSN model based on homogenous aggregates of machineries concurring for a common goal, defined as worker. Such template is then used to develop a non-linear controller to increase the power efficiency of the OSN by adjusting the frequencies of each worker, according to user-generated load of requests. Simulations are provided to show the effectiveness of the approach.