基于贝叶斯网络的通信网络故障定位方法

Fault location method of communication network based on Bayesian network

为了有效定位和修复网络故障,提出一种基于贝叶斯网络的通信网络故障定位方法。首先分三个层次构建故障传播模型,通过贝叶斯学习构建贝叶斯网络;然后将实时告警数据根据时间窗口进行分组,基于贝叶斯推理算法,使用NeticaAPI推理出可能的故障集合;最后使用评估算法对故障推理结果进行评估,进一步提高了故障定位的准确率。在故障定位时优先考虑告警级别高的告警。实验结果表明,该方法可以在一定程度上减少故障定位的时间,提高故障定位的效率和准确率。

To locate and repair network fault effectively,a communication network fault location method based on Bayesian network was proposed.Firstly,a fault propagation model was constructed in three levels and a Bayesian network was constructed by Bayesian learning.The alarms were grouped according to time window,and NeticaAPI was used to predict possible faults based on reasoning algorithm.Finally,an evaluation algorithm was used to evaluate the results,which improves the accuracy of fault location.The alarms with high priority would be considered during fault location.Experimental results show that the time of fault location is reduced;moreover the efficiency and accuracy of fault location are improved.