Energy Aware Resource Provisioning for Multi-Criteria Scheduling in Cloud Computing

Abstract

Listen

Translate article

In cloud, scheduling task requests on heterogeneous resources has proved to be a crucial research problem, particularly in terms of performance and energy consumption. The main challenge for cloud providers is to find the best tradeoff between two contradictory aims: improving user satisfaction and reducing energy consumption. Initially, to provide greater user satisfaction, the cloud computing data centers need to allocate resources according to the different users’ Quality of Service (QoS) preferences in a dynamic manner. Furthermore, to manage energy consumption on the cloud system during the execution of applications, the energy scaling decision should be made through a proper approach. In this paper, a multi-criteria scheduling algorithm based on fuzzy AHP-TOPSIS hybrid methodology is proposed to meet these requirements, which first deploys FAHP to rank and then selects the best cloud solutions according to the user’s requirements using FTOPSIS. It is then followed by effective execution of the user’s requests on available resources. Besides, an energy aware cloud resource provisioning approach based on the integration of horizontal scaling and dynamic voltage frequency scaling (DVFS) techniques is proposed to meet QoS requirements of all task requests and to manage energy consumption. The evaluation of this paper is run through different job requests while being subject to various constraints. The simulation results show that the proposed approach decreases the energy consumption and response time while enhancing both the resource utilization and user satisfaction compared with the SHARP and the BULLET algorithms in all conditions.