A Summary of Traffic Identification Method Depended on Machine Learning
Baihe Ma, Hui Zhang, Yao Guo, Zhihong Liu, Yong Zeng · 2018
Traffic classification is an invaluable tool which can category network traffics into a number of traffic classes according to various parameter settings. It can not only provide better QoS, but also protect the security of networks. In this paper, the methods of traffic classification are analyzed and summarized based on machine learning through three aspects: supervised machine learning, unsupervised machine learning, and semi-supervised machine learning. Finally, we suggest several strategies to improve the accuracy and scalability of traffic classification.