Efficient Utilization of IaaS Cloud using Adaptive Evolution Based Technique

Abstract

In the present era of technology and innovation, computing provides as a service. The service-oriented computing paradigm is a subset of utility and distributed computing paradigm. Hardware specification, platform, and application provide a service to the users in a different time zone. Resource utilization and request completion time is the prime focus of the service providers and consumers. In this work, we proposed a nature-inspired evolution and astrology science-based approach. The Big-Bang Big-Crunch optimization method based task allocation technique focuses on efficient resource utilization of IaaS (Infrastructure as a Service) cloud. We focused on three performance metrics, which may be user-oriented and service provider oriented. Performance metrics measure in terms of average resource utilization cost. Average Finish time, an average start time, and operational cost (customer-oriented) performance metrics are consumer-centric. The simulation performs using five cloud resources, six cloud configurations, and user-defined population size, and number iterations. The selection operator includes a fitness function that depends on estimated resource cost in the duration of execution. BB-BC cost-aware model provides optimal resource utilization, and minimum average finish time, and minimum average start time (ms) of cloudlets. The results exhibit that the proposed astrology, evolution based meta-heuristic approach outperforms the static (first come first serve (FCFS)), dynamic, and meta-heuristic (Genetic cost-aware, Genetic execution time aware and Particle Swarm optimization) approaches.