Multi-Objective Hybrid Optimization based Dynamic Resource Management Scheme for Cloud Computing Environments

Abstract

Nowadays, the field of Cloud Computing has particularly been becoming an emerging paradigm to deliver services over the internet. The cloud data center needs a set of virtualized resources for providing cloud infrastructure services based on user's demand. The need for optimal resource management turns into a vital issue in cloud data center to reduce the level of resource wastage and Service Level Agreement (SLA) violation. Meta heuristics algorithms are well suitable to resolve this issue which comes under NP-hard problem. The paper, proposes a Multi-Objective Hybrid fruit fly Optimization (MOHFO) based scheme for SLA aware dynamic resource management in cloud data center. The Bald Eagle Search (BES) optimization behaviour is adopted to enhance the searching ability for fruit fly optimization algorithm. The proposed scheme follows a dynamic virtual machine (VM) deployment and consolidation scheme to obtain a trade-off among SLA violation and resource wastage. The proposed scheme is simulated on Cloudsim with different data centre configurations and experimental results are evaluated based on QoS metrics. Moreover, the proposed MOHFO scheme is evaluated and compared with other optimization schemes with respect to resource wastage, energy consumption, communication cost and number of migrations performed in order to provide QoS aware optimal resource provisioning in cloud computing.