Authors : Sunita Bansal and Chittaranjan Hota
Abstract: Workflow scheduling is solved using heuristics and meta-heuristics. Heuristics are used to get an optimal solution while meta-heuristics are used to get near optimal solution. Meta-heuristics are general purpose method of solving different types of problem. This study puts Double Hybrid Multi-objective Non-Dominated Sorting Genetic Algorithm (DHNSGA-II) that gears up the convergence of the algorithm. DHNSGA-II does hybridization at two levels. At the first level it uses pre-selection operator. At the second level it uses Memetic algorithm. Pre-selection operator seeds the DHNSGA-II with the previously generated solutions. Memetic algorithm improves the current population using multi-objective local search. Apart from DHSNGA-II researchers introduced an approach to rank the Pareto frontiers because Pareto frontier has many solutions; it is nearly impossible to choose the best solution. The experimental result reveals that the proposed approach in this research performs well in optimizing the workflow scheduling jobs.
Sunita Bansal and Chittaranjan Hota, 2015. Hybrid Multi-Objective Workflow Scheduling on Utility Grids. International Journal of Soft Computing, 10: 6-18.