Scheduling Imprecise Computation Tasks with Intra-Task / Inter-Task Dependence

Abstract This work presents two heuristics to be used as admission policy in imprecise task systems. It is assumed that imprecise tasks have intra-task and inter-task dependence and a 0/1 constraint. These heuristics are supposed to be used combined with off-line schedulability tests and on-line acceptance tests already described in the literature. The objective of the admission policy is to maximize system utility through the selection of optional parts for execution. Simulation is used to analyze the proposed heuristics.