Energy-Effective Service-Oriented Cloud Resource Allocation Model Based on Workload Prediction

Abstract

The rising demands of cloud computing tend to increase the energy consumption. So, a sustainable computing environment is essential for ensuring efficient resource allocation considering the quality of service (QoS). There are many approaches in the literature employing for minimizing energy use in cloud. Predicting workload is one of the most robust and promising tasks of energy-aware cloud computing. This paper presents a service-oriented model for determining future resources requirement by predicting cloud workloads. The model incorporates several key issues alongside with load predictor to establish an energy-effective cloud environment. The workload prediction is accomplished with Multilayer Perceptron (MLP) because of its better prediction quality than the most commonly used approaches. Moreover, an implementation architecture of the proposed model is suggested to achieve the goal of this paper.