中文摘要：

本文以SNMP网络管理模型的管理信息库（MIB）为基础,在不同层次上构建了用于故障判别与定位的贝叶斯网络。对MIB变量采用自适应自回归（AAR）模型建模分析,构建与其相关协议之间的贝叶斯网络,推断协议功能是否发生异常。分析各个协议之间的功能依赖关系,构建协议间的贝叶斯网络,定位协议间的故障根源。考虑网络中故障传播构建了基于网络拓扑的贝叶斯网,定位故障根源节点。最后,对构建的模型进行了实验仿真,并分析了模型的优点和缺点。

英文摘要：

A Bayesian network is built on the basis of Management Information Base（MIB） in the SNMP network management model.The MIB variables are modeled by adaptive autoregression model,the relationship between MIB variables and related protocol is analyzed to build corresponding network to infer whether the function of this protocol is normal.ABayesian network on the basis of dependencies between protocols is built to detect the root fault among protocols.Consider the failure propagation in the network;another Bayesian network is built to locate the root node of the faults.Finally,a simulation experiment is carried out,the advantages and disadvantages of the model are analyzed.