Grid classified optimization scheduling algorithm under the limitation of cost and time

The problem of scheduling a set of independent tasks (a meta-task) with multiple QoS needs, such as time and cost, is a very important problem in grid environment because the resources are geographic distributed, heterogeneous, dynamic and owned by various organize domain. According to the cost requirement and the demand of executing time of grid task, a classified optimization-scheduling algorithm is introduced. This algorithm can well trade off the actual executing cost and its budget in grid economy model. Simulated experiments in virtual grid environment show that it has the merit that it can satisfy the demand of different grid users.