Research Article
Learning Based Genetic Algorithm for Task Graph Scheduling
- Calculate the t-level + b-level values for all the tasks; | - Considering t-level + b-level, identify all the tasks placed on the critical path; | - Repeat | - Select a task amongst the tasks ready to schedule | - If the selected task placed on the critical path Then | Assign the task to a processor that has assigned other tasks on the critical path; | Else Find a task with highest communication cost with this task and locate both of them on | the same processor; | - Until the list of tasks is empty. | - At first, each gene on the chromosome is in the boundary state or zero depth and its probability is: | |
|