A New Method for Task Scheduling in Cloud Computing by Combining PSO Algorithm and Fuzzy Logic

Scheduling tasks is one of the most critical challenges in cloud computing, directly impacting service efficiency. Task scheduling, by definition, involves allocating tasks to available resources based on their requirements. For efficient allocation of resources to users, a scheduling algorithm is necessary to determine the minimum response time, execution time, cost, and reliability. Thus, workflow scheduling is fundamentally considered an optimization problem, classified under the NP-hard category. Numerous methods have utilized metaheuristic algorithms for this purpose. This paper introduces a hybrid algorithm combining PSO (Particle Swarm Optimization) with fuzzy logic. Initially, a method for mapping resources to tasks in the cloud environment is proposed, followed by the investigation of the computation of task execution time. Subsequently, based on the swarm intelligence algorithm, an initial population of birds is created considering the number of selectable resource options and the number of tasks. By generating new generations, the birds move towards optimal points. In subsequent stages, the birds' movement speed is determined based on fuzzy logic. Finally, an evaluation of the fuzzy particle swarm optimization algorithm is conducted.

Keywords:

Cloud computing , task scheduling , PSO algorithm , fuzzy logic , optimization , ,