Emerging trends and challenges in task scheduling algorithms for cloud computing

Abstract Task scheduling is one of the important factors in cloud computing, which directly affects resource utilization, energy consumption, operational cost and overall system performance. Conventional scheduling techniques are no longer sufficient to cope with dynamic resource demands, task dependencies, and multiple conflicting objectives like minimizing makespan, reducing energy consumption, ensuring fault tolerance, and maintaining Quality of Service (QoS) due to growing complexity and heterogeneity of cloud workloads. We present an extensive survey of cloud task scheduling strategies and categorize the existing approaches into traditional, heuristic, meta-heuristic and hybrid algorithms in this paper. Each of the categories is critically analyzed based on its scheduling efficiency, scalability, resource utilization, computational complexity, and suitability for diverse cloud environments. Moreover, it analyzes major scheduling objectives such as load balancing, reliability, scalability, energy efficiency, and cost minimization, considering practical constraints such as execution deadlines, budget limits, security requirements, and resource availability. We also investigate the adaptability of static, dynamic, artificial intelligence (AI)-driven and prediction-based scheduling techniques for an effective task-to-resource mapping under different workload conditions. Moreover, a comparative study among the popular Cloud simulation frameworks such as CloudSim, WorkflowSim, GreenCloud, etc. is also presented in the paper, highlighting the pros and cons of these frameworks for evaluating scheduling algorithms in a controlled experimental environment. Finally, future research directions such as AI-enabled scheduling, energy-aware resource management, and task scheduling in cloud, edge, and fog computing environments are discussed. This review presents useful insights into the state-of-the-art cloud task scheduling, and highlights the promising opportunities for designing intelligent, scalable and energy-efficient scheduling solutions for the next generation of cloud computing systems.