In view of such problems as temporary loops

blackholes

violations of access control policy and so on during state transitions in OpenFlow networks

we proposed a consistent rule update scheme based on heuristic scheduling.First

we divided operations of an update into some sub-updates and used parallel network property verification technique to construct a dependency graph for each sub-update.Second

we aggregated several update operations of the same sub-update to reduce complexity of scheduling.Then

we designed heuristic update scheduling algorithm.By adopting an alternative strategy for additions and deletions of rules

it could reduce storage cost in switches.Furthermore

by establishing a mechanism to update and monitor concurrently

it could improve efficiency of rule update.Simulation experiments on the updating time cost and the rule storage cost of switches during updating process are conducted

which verify the effectiveness of our scheme.