Hybrid technique for optimal task scheduling in cloud computing environments

Abstract

Since cloud computing has an abundance of users, the system has to execute a wide range of tasks. Task scheduler methods that are both robust and efficient while delivering the best outcomes are required. The task volume and runtime in the cloud vary rapidly, making task assessment and resource mapping difficult. Security issues, communication delays, and data loss are substantial barriers to scheduling. Furthermore, optimization techniques can be utilized to reduce load and assign tasks so that the user can finish tasks faster. This paper offers a hybrid job scheduling technique for cloud computing using adaptive particle swarm optimization and ant colony optimization particle swarm optimization-ant colony optimization (adaptive PSO-ACO). After rapidly finding the initial solution via particle swarm optimization, the ant colony optimization approach establishes its first pheromone distribution. The suggested hybrid algorithm is compared to standalone PSO and ACO algorithms. Compared to ACO, the percentage decrease is 7.9%. Hybrid method has the lowest total cost, 55% less compared to PSO. Tasks vary when virtual machines (VMs) are constant and VMs vary when tasks are constant. Parameters like final cost, makespan, fitness value, computation time and weighted time are assessed to evaluate the performance of the hybrid task scheduling algorithm.