Optimizing cloud computing: An improved algorithm-driven scheduling framework for independent tasks

Cloud computing (CC) has been recognized as a foundational element of contemporary computing infrastructure, enabling cost-effective utilization of resources and high-quality service delivery. Nevertheless, task scheduling remains a critical challenge due to its NP-hard complexity, which adversely affects execution efficiency and resource planning. As scheduling plays a pivotal role in minimizing key performance metrics—such as cost, energy consumption, and makespan—a novel approach based on the Dynamic Earthworm Optimization (DEWO) algorithm has been proposed for scheduling independent tasks with mixed properties in cloud environments. The DEWO algorithm was designed to enhance scheduling efficiency by dynamically adapting its parameters in accordance with task characteristics and system conditions. Its effectiveness was evaluated using the CloudSim toolkit, a widely adopted simulation platform in cloud computing research. A range of experimental scenarios was developed to assess DEWO’s performance under varying workloads and task allocations. Key performance indicators—including cost, energy usage, and makespan—were analyzed to determine the method’s efficiency, and statistical tests were conducted to validate the reliability of the results. The DEWO algorithm demonstrated significant improvements in throughput and reductions in total completion time compared to existing optimization techniques. The experimental outcomes confirmed that DEWO is capable of efficiently scheduling tasks with diverse computational demands, thereby enabling optimal resource utilization. Across a variety of workload conditions, DEWO consistently outperformed competing methods, highlighting its robustness and adaptability. Through an adaptive mechanism, a balance between exploration and exploitation was achieved, resulting in more effective scheduling decisions. These findings indicate that DEWO holds strong potential as an effective solution for dynamic cloud task scheduling, due to its ability to reduce execution time and cost while maximizing the use of available resources.