Network traffic protocol recognition method based on deep learning

Abstract

The invention discloses a network traffic protocol recognition method based on deep learning. The similarity of network traffic data and images is utilized, the work of traffic feature value selectionand extraction is skipped and the network traffic data directly act as the input of a convolutional neural network to perform supervised learning and train a network traffic protocol recognition model so that the function of network traffic protocol recognition can be realized. According to the method, only a network traffic protocol sample to be recognized is provided for training of the convolutional neural network so that the features favorable for the classification task can be automatically extracted without spending energy on protocol feature extraction and selection; and the method haslearning and extending capacity and can be used for recognizing the new unknown protocol and can be extended to be applied to recognition of the network traffic including malicious codes and recognition of the corresponding program traffic.

Description

technical field [0001] The invention relates to the technical field of network security, in particular to a network traffic protocol identification method based on deep learning. Background technique [0002] A network protocol is a collection of rules, standards or conventions established for data exchange in computer networks. The identification and analysis of network protocols is the basis of network security and is of great significance to network supervision, anomaly detection and maintenance of network security. Traditional network traffic protocol identification methods mainly include: port-based network traffic protocol identification, network traffic protocol identification based on deep packet inspection (DPI), and network traffic protocol identification based on machine learning methods. [0003] Port-based network traffic protocol identification method: In a traditional network environment, various network applications follow established rules and use specific ...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to overcome the deficiencies in the prior art, provide a network traffic protocol identification method based on deep learning, and solve the technology in the prior art that the traffic protocol method has a limited scope of application, difficult identification, and high requirements for the network environment question

Images