Research on intelligent model-based core network anomaly detection technology
The detection of core network KPI anomalies is crucial for ensuring the stable operation of the network.To promote the development of intelligent network operations and maintenance,this paper proposes a core network KPI anomaly detection method based on an intelligent model.This method first uses a single-layer decision tree to process KPI data,dividing the data into two categories:boundary-type anomalies and non-boundary-type anomalies.Next,for the boundary-type anomaly KPI data,three models—single-layer decision tree,Naive Bayes,and K-Nearest Neighbors(KNN)—are used for evaluation.The results of these three models are then integrated using the VOTING method to further improve the accuracy and comprehensiveness of anomaly detection.Experimental results demonstrate that this method achieves good performance in detecting core network KPI anomalies.
万方数据
维普
