Level based task duplication strategy to minimize the job turnaround time

One of the major addressable issues in Parallel and Distributed Systems is scheduling i.e. mapping of tasks on processing nodes in order to optimize the QoS parameters. Task Duplication is an effective approach to minimize turnaround time and communication overheads and to improve system robustness. Task duplication effectively ensures redundantly execution of some task on which some other task critically depends. The task scheduling with duplication and without duplication is known to be an NP-complete problem. This work proposes a static scheduling based on module dependence degree and task duplication in order to minimize the turnaround time of the job.