A look-ahead strategy-based method for scheduling multiprocessor tasks on two dedicated processors

In this paper, we investigate the use of a look-ahead strategy combined with path-relinking for tackling problem scheduling multiprocessor tasks on two dedicated processors. An instance is composed set divided into three subsets and processors, where some can be executed either one processor or The goal to schedule all such that execution last assigned task minimized. First, proposed method starts greedy feasible solution built by applying knapsack rule related predefined sequence. Second, series local operators are added in order drive search process around neighborhoods. Third, first diversification based upon drop rebuild employed. second diversification/intensification used highlighting performance method; it incorporates path-relinking. Finally, experimentally analyzed benchmark instances literature, its provided results compared those achieved more recent methods available literature. Its behavior also evaluated providing statistical analysis using both Sign test Wilcoxon signed-rank test.