An Optimized Hybrid Approach for Load Balancing and Task Scheduling in Cloud Computing Environment

Abstract

– Cloud computing has evolved as a significant platform to provide scalable and as-needed services, but efficient scheduling of tasks and load balancing is still difficult due to the heterogeneous virtual machines and dynamic workloads. This paper presents a hybrid model, O-MCTSALP, which is optimized to schedule tasks and balance their loads in cloud computing systems. The suggested approach is a hybrid optimization mechanism with Minimum Completion Time (MCT)-based scheduling to increase the allocation of tasks to virtual machine and overall system performance. The model was tested in simulated cloud environment in MATLAB and compared with MCTFFA and HFFSSA, PSO optimization and EASA-MORU at workloads of 100, 500, 1000, 2000 and 5000 tasks. The experimental findings indicate that O-MCTSALP had the lowest makespan of all the workloads that were tested with the values being 95.9483, 158.4058, 316.3737, 557.8421, and 1657.1 respectively. The approach also yielded good resource utilization of medium and large workloads to 0.9516 at 500 tasks, 0.9835 at 1000 tasks, 0.9882 at 2000 tasks, and 0.9968 at 5000 tasks. Moreover, O-MCTSALP had lower values of load imbalance of 53, 56, 54, 59 and 55, and waiting time of 8.9364 and 8.2316 at 500 and 2000 tasks respectively. The model also enhanced throughput to 3.1565, 3.1608, 3.5852 and 3.0173 of 500, 1000, 2000 and 5000 tasks respectively. These results show that O-MCTSALP is a promising and scalable method of enhancing the efficiency of scheduling, resource consumption, and task distribution in dynamic cloud computing systems.