Network Traffic Identification Based On Machine Learning and Deep Packet Inspection

Abstract

Accurate network traffic identification is an important basic for network traffic monitoring and data analysis and is the key to improve the quality of user service. In this project, through the analysis of two network traffic identification methods based on machine learning and deep packet inspection, a network traffic identification method based on machine learning and deep packet inspection is proposed. The deep packet inspection based on the feature library RuleLib, conducts in depth analysis of data traffic through pattern matching and identifies specific application traffic. Machine learning method is used to assist in identifying network traffic with encryption and unknown features, which makes up for the disadvantage of deep packet inspection that cannot identify new application and encrypted traffic. Experiments show that this method can improve the identification rate of network traffic.