Research Article
6G Network Traffic Intrusion Detection Using Multiresolution Auto-encoder and Feature Matching Discriminator
@INPROCEEDINGS{10.1007/978-3-031-36011-4_18, author={Yuhai Li and Yuxin Sun and Dong He and Liang Xi}, title={6G Network Traffic Intrusion Detection Using Multiresolution Auto-encoder and Feature Matching Discriminator}, proceedings={6GN for Future Wireless Networks. 5th EAI International Conference, 6GN 2022, Harbin, China, December 17-18, 2022, Proceedings, Part I}, proceedings_a={6GN}, year={2023}, month={7}, keywords={6G intrusion detection autoencoder generative adversarial network}, doi={10.1007/978-3-031-36011-4_18} }- Yuhai Li
Yuxin Sun
Dong He
Liang Xi
Year: 2023
6G Network Traffic Intrusion Detection Using Multiresolution Auto-encoder and Feature Matching Discriminator
6GN
Springer
DOI: 10.1007/978-3-031-36011-4_18
Abstract
With the development of 6G technology, security and privacy have become extremely important in the face of larger network traffic bandwidth. An effective intrusion detection system can deal with the network attacks. Deep learning has been developed in the field of intrusion detection, which can identify normal and abnormal traffic. However, existing methods cannot guarantee good performance in accuracy and efficiency. In this paper, based on the autoencoder and generative adversarial network, the multiresolution autoencoder is adopted in the network traffic feature extraction, which can obtain different encoding lengths and guarantee better data reconstruction. In addition, we add an extra feature matching loss to encourage the discriminator to get more discriminative information from the reconstructed samples. Our experimental results on the CIC-IDS2018 dataset indicates that compared with autoencoder and generative adversarial network, our model can effectively improve the detection accuracy and can be applied to 6G network traffic security detection.
