MFGMTS: Epsilon Constraint-Based Modified Fractional Grey Wolf Optimizer for Multi-Objective Task Scheduling in Cloud Computing

Optimization of computing resources in cloud computing requires a scheduling algorithm so that the user-requested tasks can be scheduled effectively. In addition to the efficiency, the adopted task scheduling algorithms must meet the user requirements. Although there are many algorithms for task scheduling, the algorithms that define multiple objectives with considered trade-off are rare. This paper proposes a multi-objective optimization algorithm, Modified Fractional Grey Wolf Optimizer for Multi-Objective Task Scheduling (MFGMTS) in cloud computing environment. The objectives, execution time, execution cost, communication time, communication cost, energy consumption, and resource utilization are computed using epsilon-constraint and penalty cost function. This newly considered constraint minimizes the fitness function, to provide optimal task scheduling. The algorithm is motivated by Fractional Grey-Wolf Optimization (FGWO) with a modification in the position update, where an additional term is incorporated using the combination of alpha and beta solutions. The algorithm is compared with the existing Particle Swarm Optimization, Genetic Algorithm (GA), Grey Wolf Optimizer, and FGWO to analyze the performance efficiency. It can attain minimum values of 0.186243, 0.174782, 0.016045, 0.087023, 0.012259, and 0.564528, regarding execution time, communication time, execution cost, communication cost, energy consumption, and resource utilization.