Multi-objective task scheduling evolutionary algorithm in mobile cloud computing

The multi-objective optimization problem of tasks scheduling in mobile cloud is studied, a multi-objective scheduling evolutionary algorithm MTSEA is proposed. The multi-objective opitization model of tasks scheduling with the constraints of deadline, budget and energy, which introduces the execution makespan, execution cost and execution energy consumption as tre-objective optimization. An evolationary algorithm is presented to solve this tre-objective optimization problem with conflict. Our algorithm focuses to introduce three population individual with highest efficiency, minimal cost and highest energy-efficiency in the population initialization instead of randomly generated particles. And, our algorithm uses the evolutionary iteration of the crossover and mutation operation among population individuals. Finally the algorithm can generate the scheduling solution set meeting Pareto optimal by the form of the not-dominant order. Some simulation experiments are constructed and the performance comparsion are performed with other two kinds of multi-objective scheduling algorithms. The results show that MTSEA performs better on the convergence of scheduling solutions and the solution space distance and its distribution.