A graphical user interface for the analysis of task graphs

This thesis describes a flexible graphical user interface (GUI) tool designed and developed for the analysis of task graphs. The interface tool takes the task graph along with parameters such as execution times of the tasks, number of processors, and communication costs, as input and gives the average completion time and the probability distribution of the completion times for the task graph as output, using a numerical analysis tool and a simulation analysis tool. The interface provides access to heuristic algorithms to generate an allocation matrix that maps tasks into processors. The solution tools uses this matrix along with the parameters to generate results. The GUI is developed in such a way that the user can configure it by adding new heuristics or any other solution scheme.