Cloud Performance Evaluation: Hybrid Load Balancing Model Based on Modified Particle Swarm Optimization and Improved Metaheuristic Firefly Algorithms

Abstract

Cloud computing is a group of devices that have been interconnected and vary across pc to network. This one promises consumer applications an immense amount of memory and enhanced processing capabilities through the network. Inside a cloud framework, the load-balancing strategy can then be used to accomplish the purpose of the utilization of optimal resources. The load-balancing strategy transfers consumer assignments to that of the appropriate Virtual Machines and achieves metrics of just the QoS standard of service. The Optimization techniques may be used to overcome complex problems with non-deterministic polynomials, like those of scheduling tasks. The whole research describes a hybrid load balancing framework centered onto an optimization of modified particle swarm with optimized metaheuristic firefly algorithms to boost cloud efficiency. The firefly approach will be used in the proposed hybrid approaches which reduce the searching range towards optimal responses, as well as the strategy of enhancing the particle swarm is used to determine the optimal solution. The whole proposed approach is centered on some kind of predictive load distribution which primarily offers resources provisioning as well as offers a load balancing model for the implementation of workflows which enhance the utilization of uniformly load distribution VMs. The whole research has indeed provided certain measures for evaluating the performance of just the hybrid approach being suggested. The findings demonstrated improved efficiency than comparable approaches, and also adaptive actions in high load avoidance by optimization of multiple objectives.

Keywords: Cloud load balancing, Modified particle sward optimization, Improve Metaheuristic Firefly, Virtual Machines, Workflow scheduling