Cloud computing has emerged as a cornerstone of digital transformation, offering scalable, flexible, and costeffective infrastructure to meet the growing demands for data processing and IT services. By enabling on-demand access to virtualized resources, this technology has fundamentally transformed the manner in which organizations consume and manage computing services. However, as cloud architectures grow in complexity, efficient resource management, particularly task scheduling, becomes a critical challenge that requires intelligent and innovative solutions. This study addresses multi-objective optimization for task scheduling in cloud environments using the Squirrel Search Algorithm (SSA), drawing inspiration from the dynamic foraging behavior of flying squirrels. The proposed approach aims to optimize conflicting scheduling metrics, including execution time, cost, and energy consumption. In order to validate the effectiveness of the proposed technique, a comparative analysis was conducted against two established optimization techniques: Particle Swarm Optimization (PSO) and Non-dominated Sorting Genetic Algorithm II (NSGA-II). The experimental results, obtained through the CloudSim simulation framework across a range of test-bed scenarios and synthetic workload traces, demonstrate the effectiveness of the SSA-based scheduler in terms of scheduling efficiency, energy awareness, and cost-effectiveness.
In real-world decision-making scenarios, particularly within cloud computing environments, multiple stakeholders with divergent objectives must be considered. In Infrastructure as a Service (IaaS) models, the task scheduling process constitutes a challenging multi-objective optimization problem due to the heterogeneity and dynamic behavior of virtualized resources. This study introduces a hybrid decision-making framework that combines the Shuffled Frog Leaping Algorithm (SFLA), a swarm intelligence-based metaheuristic, with four established Multi-Criteria Decision-Making (MCDM) methods: MAIRCA, EAMR, MARCOS, and WSM. SFLA is used to generate a diverse set of scheduling alternatives, which are subsequently evaluated with respect to three conflicting objectives: minimizing makespan, reducing execution cost, and maximizing resource utilization. Simulation experiments conducted with the CloudSim toolkit demonstrate the effectiveness of the proposed approach and reveal a fundamental trade-off between economic efficiency and infrastructure performance, providing valuable insights for cloud service providers seeking to balance user satisfaction and operational optimization.