With the rapid development of Internet technology, people gradually rely on intelligent devices for information exchange and communication through the network. With the popularity of smart connected devices, security issues are growing, including malicious traffic attacks. Therefore, the design of an efficient malicious traffic detection system is of great significance. This paper makes two improvements based on the traditional malicious traffic classification model for malicious traffic detection. One is to adopt a brand new combined dataset named Composed Encrypted Malicious Traffic Dataset and conduct a correlation coefficient matrix analysis on the data set. The second is to design an improved SVM model based on the gray wolf optimization algorithm. We abandon the idea of relying on expert experience to set parameters and optimize parameters through intelligent algorithms. Through six groups of comparative tests with GWO-SVM, Xgboost, RF, SVM,GA, GA-BP, the classification effect of the improved model is the best, and the classification accuracy is 0.99625, the standard deviation for cross-validation is 0.01.
Malicious traffic detection based on GWO-SVM model
2022-10-12
1305700 byte
Conference paper
Electronic Resource
English
MALICIOUS TRAFFIC DETECTION WITH ANOMALY DETECTION MODELING
European Patent Office | 2022
|MALICIOUS TRAFFIC DETECTION WITH ANOMALY DETECTION MODELING
European Patent Office | 2022
|Research on Malicious TLS Traffic Detection Based on Spatiotemporal Feature Fusion
Springer Verlag | 2022
|