Multi-Objective Optimization for Dynamic Resource Provisioning in a Multi-Cloud Environment using Lion Optimization Algorithm
Citations Over TimeTop 12% of 2020 papers
Abstract
Cloud computing offers surfeit of services like storage, computing power, run-time environment, network etc. that everyone is accustomed to use it in day-to-day lives. In cloud computing, resources need to be dynamically provisioned on a metered basis. Quality of Services(QoS) is promised like performance, scalability, efficiency, fault tolerance, availability, reliability, throughput, and so on. Several meta-heuristic nature-inspired optimization algorithms like Particle Swarm Optimization(PSO), Ant Colony Optimization(ACO) etc. deployed to meet Service Level Agreement parameters like minimum downtime and low latency, but still have challenges in dynamically allocating resources. To overcome the above stated challenges, a new dynamic resource provisioning technique in a multi-cloud environment is proposed that uses Lion Optimization Algorithm(LOA) wherein characteristics of nomad and pride lion groups are taken into account. The multi-cloud environment provides the organization or customer to choose a provider that meets the specific requirements. As compared to PSO, this approach achieved better results while optimizing multiple objectives like completion time, average response time, makespan, cost, and average resource utilization. This study proves that the completion time and cost for LOA has outperformed when compared to PSO for a given number of tasks. The makespan and average response time for LOA improves slowly with more number of tasks as compared to PSO.
Related Papers
- → Predicting Cloud Resource Provisioning using Machine Learning Techniques(2023)19 cited
- → An Adaptive Approach to Resource Provisioning in PaaS(2011)1 cited
- → Diagnosing Memory Provisioning in IaaS Clouds(2013)
- Application of Expected Loss Provisioning and Dynamic Provisioning(2011)