Fault Diagnosis Algorithm Based on Service Characteristics Under Software Defined Network Slicing

In order to solve the problem of low accuracy of fault diagnosis algorithms brought by network dynamics, this paper proposes a fault diagnosis algorithm based on service characteristics under software defined network slicing. In order to reduce the problem of inaccurate symptom information caused by network dynamics, the credibility of symptoms is calculated based on the alternative probabilistic characteristics of network nodes, and the symptom information is corrected. The node importance is analyzed from the two dimensions of node centrality and number of links. Based on the node importance and symptom information, the reliability of the node failure is ranked. Finally, based on the maximum coverage algorithm, the optimal set of suspected faults is selected from the set of suspected faults as the final set of faults. The experiment compares the algorithm in this paper with the existing algorithm, and verifies that the algorithm in this paper effectively improves the accuracy of fault diagnosis.