A Trustworthy Network Fault Diagnosis Approach

Trustworthy network fault diagnosis approach is one critical management item to enhance network trustworthiness. Aiming at gaining highly trustworthy of fault diagnosis in Internet, we present a trustworthy fault diagnosis approach based on integration of Artificial Neural Network and Rule-based Reasoning. Supported by the topology of hierarchical and distributed in multi-domains, reasoning rule matrix and its operation are studied to acquire parallel reasoning capability. Moreover, the quantitative trustworthy degree is defined and information entropy is applied to define threshold function marked on arcs and nodes in the Artificial Neural Network. Our approach possesses higher parallel capability guaranteeing by matrix operation and trustworthiness by trustworthy degree definition and calculation using Artificial Neural Network. Further, it is general so that it can be transplanted into various application fields.