摘要
由于导致网络运维场景异常的因素较为复杂,直接利用表观数据对异常状态进行检测,误差相对较大,为此,提出基于故障树的IMS网络运维场景异常检测算法。利用Fusell-Vesely算法,以运维场景的比特率为参数,按照逐级向下分解的方式计算故障树的最小割集,通过顶事件下属事件之间的逻辑关系实现对网络运维场景异常因素的全覆盖,在获取待检测场景的比特率数据后,借助故障树逐级判断数据对应的事件,并得到最终的状态判定结果。测试结果中,设计算法对不同程度的IMS网络运维场景异常检测结果误差稳定在1.0%以内。
Due to the complexity of the factors that lead to the anomaly of network operation and maintenance scenarios, the anomaly state is detected directly by using the apparent data, and the error is relatively large. Therefore, an IMS network operation and maintenance scenario anomaly detection algorithm based on fault tree is proposed. The Fusell Vesely algorithm is used to calculate the minimum cut set of the fault tree with the bit rate of the operation and maintenance scenario as the parameter, and the logical relationship between the subordinate events of the top event is used to achieve full coverage of the abnormal factors of the network operation and maintenance scenario. After obtaining the bit rate data of the scenario to be detected, the event corresponding to the data is judged level by level with the help of the fault tree, and the final state determination result is obtained. In the test results, the error of the designed algorithm in anomaly detection of IMS network operation and maintenance scenarios of different degrees is stable within 1.0%.
出处
《现代传输》
2023年第1期61-64,共4页
Modern Transmission
