Adversarial Sample Generation Based on Model Simulation Analysis in Intrusion Detection Systems

The explosive development of artificial intelligence technology is profoundly affecting the strategic landscape of cyberspace security, demonstrating enormous potential in the field of intrusion detection. Recent research has found that machine learning models have serious vulnerabilities, and adversarial samples derived from this vulnerability can significantly reduce the accuracy of model detection by adding slight perturbations to the original samples. In our article, we propose a novel method called adversarial sample generation based on model simulation that quickly generates adversarial samples in a short period of time and improves the model’s generalization and robustness after adversarial training. Extensive experiments on the CICIDS-2017 dataset demonstrated that the method consistently outperforms other current research methods.