A 5G Slice Traffic Anomaly Detection Method Based on Convolution Neural Network

As the network security threats caused by the Internet of Things (IoT) continue to increase, the attack surface continues to expand and the network heterogeneity increases, and the use of virtualization technology and distributed architecture increases. Protecting the future network will become a challenging field. This paper proposes the approach of converting 5G slice traffic into images for analysis by a CNN (Convolutional Neural Network) model, and uses NAS (Neural Architecture Search) to detect anomalous network traffic. In order to improve the accuracy and efficiency of the intrusion detection system, this paper uses SDS(Software Defined Security) as a means to design a scalable, high-efficiency 5G traffic defense system. In the multiclass classification, the accuracy of CNN model is 94.9\%. The results from this method are promising as the model has performed in multiclass classification with an accuracy of 96.4\%. Furthermore, we suggest the optimal CNN design for the better performance through numerous experiments.