Work

Malicious traffic detection method based on field feature deep learning

Year: 2025

Type: conference-paper

Abstract: With the rapid development of cloud computing services and security technologies, more and more enterprises are embracing deep binding of business with cloud services, but at the same time, malicious attacks targeting cloud services are gradually becoming rampant. Since the traffic in the cloud service log contains more comprehensive and rich semantic information, malicious traffic detection can be seen as the anomalous text classification task. Therefore, this paper proposes a novel detection method based on deep learning of field features. Firstly, the word vectorization preprocessing is carried out through field selection, decoding, and character encoding, and then the improved TextCNN (Convolutional Neural Networks for Sentence Classification) is used to extract the field features from multiple dimensions. Finally, the classification is carried out. Experimental results show that, compared with the classical models SVM and RF (Random Forest), our proposed method achieves better classification results in malicious cloud service traffic detection tasks.

Source: International Conference on Computer Application and Information Security (ICCAIS)

DOI: 10.1117/12.3062181

Authors: Zhouguo Chen, Chen Deng, Xinze Li, Xiang Gao, Jiayue Mao, Hangyu Hu

Institutions: China Electronics Technology Group Corporation, Southeast University, University of Electronic Science and Technology of China

Language: English

Topic: Network Security and Intrusion Detection

Subfield: Computer Networks and Communications

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funders: National Natural Science Foundation of China, Fundamental Research Funds for the Central Universities

Award: G552993833