A new heuristic for minimizing schedule length in heterogeneous computing systems

Heterogeneous environments includes resources with diverse capabilities necessitates efficient task-to-processor assignment for accomplishing high performance. In the proposed work a new heuristic approach for scheduling the tasks independent of their levels in the precedence constrained task graph is detailed. Schedules with shorter span are achieved by reducing the start time of the independent tasks. The proposed algorithm is compared with the available literature and an improvement in the schedule length is obtained.