Time and cost trade-off using multi-objective task scheduling in utility grids

As the applications of utility grids and market-based grids expand, the need for scheduling algorithms which consider the users´ often conflicting preferences also increases. The users want their applications executed in the most possible economical way within minimum time. This paper proposes a multi-objective task scheduling in utility grids with two important objectives: cost and makespan. The proposed algorithm applies Maximin fitness function along with genetic operators and also uses two heuristics for neighborhood searches. Simulation results with GridSim on twelve different instances show that in most cases, the proposed method puts forward a better diversity and quality scheduling options comparing to other existing algorithms.