Optimization of Network Traffic Anomaly Detection Based on K-Means and Regression Analysis

With the rapid development of 5G network technology in China, the detection of network traffic, especially mobile network traffic anomalies, has become a key work in the field of modern network security. This paper proposes a method for network traffic anomaly alarm detection based on the algorithm of K-means clustering and regression analysis. Based on the collection of multi-modal sensors in the network, combined with the threshold value after kmeans clustering to trigger the relevant management, real-time monitoring, classification and early warning management of network traffic is realized. Due to the k-means clustering algorithm, it can effectively identify the special situation of each parameter of the traffic, and use regression analysis to effectively predict and assess the relevant degree of abnormality, and introduce the relevant managers to intervene, combined with different clustering results to trigger different network management strategies, which improves the management process of the quality of network security.