Flow Table Update Strategy Based On SDN

SDN(Software Defined Network)is a new network framework and technology based on software,it's significant features is that it had loose coupled controlling plane and data plane,besides it supported centralized network controlling state,and it realized the opacity from the underlying network to the upper application.For the problem of the low storage efficiency of the OpenFlow switch and the safety of data transferring under the circumstance of safely transmission.Main work of this paper as follows:This paper proposed a matching mechanism of multistage flowing tables,and the paper made a multilevel segmentation of the matching flowing table in the OpenFlow switch.The paper proposed a storage strategy of multistage flow tables fine-grained.The paper analyzed the refreshing complexity of flowing tables in TCAM(Ternary Content Addressable Memory).The simulation results showed that fine-grained division could increase the storage efficiency of TCAM,besides it did not increase the complexity of the refreshing of flowing tables.The paper also proposed ADCMC(Average Delay Constraints Minimum Cost)algorithm of safe transmission path of SDN.The paper absorbed the transmittal link cost in the topology into the equitable delay function to get the optimal solution according to the thoughts of Lagrangian relaxation algorithm,and solved the time delay and safety problem of point to point transmission.The simulation results showed that,when the topological structure was very complex and had an amounts of links and node,ADCMC algorithm is better than the Lagrangian relaxation algorithm of the average costs.