A CE-GAN based approach to address data imbalance in network intrusion detection systems.

As network intrusion behaviors become increasingly complex, traditional intrusion detection systems face limitations, especially with data imbalance. To address this, we introduce the Nash equilibrium concept from game theory into classifier ensemble optimization, enhancing robustness in multi-class classification tasks. Additionally, we propose a network intrusion detection system based on a Conditional Generative Adversarial Network with Conditional Aggregation Encoder-Decoder Structure (CE-GAN) with a conditional aggregation encoder-decoder structure to mitigate data imbalance and improve classifier performance. The model incorporates a composite loss function to maintain both the authenticity and diversity of generated samples. Experiments on the NSL-KDD and UNSW-NB15 datasets show that CE-GAN effectively augments rare data samples, significantly improving classification metrics for imbalanced datasets, thus providing a superior solution to this challenge in network intrusion detection.