GAN-based intrusion detection model using MLP encoder

作者：Xiao Zhang, Xiaoming Chen, Yuxiong He, Youhuai Wang, Yong Cai, Huixun Li · 年份：2023 · DOI：10.1117/12.2685912 · 被引用次数：1 · 研究领域：Network Security and Intrusion Detection、Anomaly Detection Techniques and Applications、Advanced Malware Detection Techniques

As the number of network attacks continues to rise, effective intrusion detection systems (IDSs) have become crucial for maintaining network security. Traditional machine learning (ML) methods have achieved good results in intrusion detection, but are often limited by insufficient training data and vulnerability to adversarial attacks. To address these limitations, we propose a neural network model based on Generative Adversarial Networks (GAN). GAN have achieved remarkable success in detecting such adversarial attacks, mainly due to the adversarial training of generators and discriminators in an attempt to bypass each other and thereby improve their own capabilities and accuracy. Furthermore, the generator of GANs can generate fake samples that are similar to real data, thereby alleviating the problem of insufficient training data. Our experimental results demonstrate that our proposed IDS model effectively detects network attacks with high accuracy. We believe that this approach represents a promising direction for future research in the field of network security.