Can intelligent optimisation techniques improve computing job scheduling in a Grid environment? review, problem and proposal

In the existing Grid scheduling literature, the reported methods and strategies are mostly related to high-level schedulers such as global schedulers, external schedulers, data schedulers, and cluster schedulers. Although a number of these have previously considered job scheduling, thus far only relatively simple queue-based policies such as First In First Out (FIFO) have been considered for local job scheduling within Grid contexts. Our initial research shows that it is worth investigating the potential impact on the performance of the Grid when intelligent optimisation techniques are applied to local scheduling policies. The research problem is defined, and a basic research methodology with a detailed roadmap is presented. This paper forms a proposal with the intention of exchanging ideas and seeking potential collaborators.