Design And Implementation Of Flow Rules' Inconsistent Detection And Resolve Algorithm

Software-defined network(SDN)is a new type of network architecture that improves the programmability and flexibility of the network by separating data and control.However,there are also many security issues.For example,an attacker can issue conflicting flow rules,changing the processing of forwarding devices.The existing research focuses on the detection and resolution of conflict relationships between flow rules,lacking of detection and resolution of other types of flow rule relationships.In addition,none of them is implemented on ONOS controllers.In order to ensure the safe and efficient execution of flow rules,this paper designs and implements a detection and resolution algorithm for inconsistent flow rules.The content of the dissertation includes the following four aspects.1 Process of verifying flow rules' validity and sender's priority is designed.In combination with md5 encryption algorithm and digital signature technology,the transmitter encrypts flow rules,while receiver end decrypt flow rules,to verify flow rules legitimacy and determine the identity of the sender of the flow rule;2 Detection algorithm for inconsistent flow rule is designed.According to the relationship between the matching domains,detect flow rules' relationship among the three type including include,mergeable,and conflict;3 Solving algorithm for inconsistent flow rule is designed.Based on the inconsistency relationship between the new flow rule and the original flow rule and its sender's priority relationship,the flow rule is processed,including the treatment of ignoring the new flow rule or replacing the original flow rule with the new flow rule;4 The rule inconsistency detection and resolution algorithm on the ONOS controller is implemented.Furthermore,function verification and performance testing are performed respectively.It is proved that solving the inconsistency of flow rules can improve the network performance by comparing experiments.