Two Novel Genetic Operators for Task Matching and Scheduling in Heterogeneous Computing Environments

Techniques for task matching and scheduling play a crucial role in harnessing the computing resources of a heterogeneous computing environment that has become increasingly ubiquitous today. In this paper, we therefore propose two sophisticated operators used in genetic algorithms (GAs) and demonstrate their effectiveness to the task-matching and -scheduling problem. These two genetic operators, namely the topological-ordered crossover (TOX) and the priority-guided mutation (PGM), incorporate the knowledge of problem characteristics to improve the solution quality obtained. On the basis of the problem-specific knowledge, moreover, a schedule generated by the TOX operator is guaranteed to be valid. For the sake of avoiding early search stagnation, the PGM operator also integrates the concepts of simulated annealing (SA). Performance of the proposed approach is demonstrated by comparing it against other existing scheduling techniques in terms of overall schedule length of randomly generated problem instances. Experimental results indicate that the proposed approach is a significant improvement compared with the previous attempts in solving the task-matching and -scheduling problem.