A Multi-objective Scheduling Algorithm of DAG Cloud Tasks

A DAG tasks scheduling algorithm FDS-PSO of discrete particle swarm optimization based on the fuzzy dominance sort is presented. First, the multi-objective optimization model of workflow scheduling under bi-QoS with the budget and deadline constraint is established, which introduces three optimizaiton objectives, including the execution makespan, the execution cost and the scheduling reliability. Second, a discrete particle swarm optimizaiton algorithm is designed to solve this three conflicting objectives optimization. FDS-PSO can quantify the relative fitness of multi-objective solutions through the fuzzy dominance sorting and obtain the solutions set of workflow scheduling satisfying Pareto optimal. Finally, through the simulation experiments of generated randomly workflow, compared with the same types of intelligent swarm multi-objective scheduling algorithms.Experimental results show that the scheduling solution obtained by FDS-PSO not noly has better convergence, but has better and uniform spacing distribution among solutions, which can better achieve multi-objective trade-off optimization.