Enhanced Task Scheduling in IaaS Cloud Environments Using Elitism-Based Genetic Algorithms

Cloud computing (CC) is a modern commercial model that enables customers to acquire large amounts of virtual resources on demand. Among the various service models in CC, Infrastructure as a Service (IaaS) provides Virtual Machines (VMs) and data centers. Efficient task scheduling, which maps cloud tasks to VMs, is key to optimizing data center performance and reducing energy consumption. Given the heterogeneous nature and computational intensity of these tasks, meta-heuristic methods are often employed for scheduling. This research proposes an enhanced Genetic Algorithm (GA) that integrates an Elitism-Based strategy with Conditional Parameter Tuning to improve convergence speed and solution quality. The elitism approach preserves top-performing solutions across generations, while conditional parameter tuning dynamically adjusts algorithm parameters based on population diversity and fitness levels. Experimental evaluations on Amazon EC2 show that the proposed method significantly outperforms traditional approaches in task completion time, resource utilization, and convergence efficiency. The results demonstrate the effectiveness of combining elitism with adaptive strategies to create a scalable, robust solution for task scheduling in high-demand cloud environments.