Search papers, labs, and topics across Lattice.
NetDiffuser is introduced as a novel framework leveraging diffusion models to generate natural adversarial examples (NAEs) capable of deceiving deep learning-based Network Intrusion Detection Systems (NIDS). The framework employs a feature categorization algorithm to identify independent network traffic features for perturbation, followed by a diffusion model to inject semantically consistent perturbations. Experiments on benchmark NIDS datasets demonstrate that NetDiffuser achieves up to 29.93% higher attack success rates and significantly reduces AE detection performance compared to existing methods.
Diffusion models can craft network attack traffic that's nearly undetectable to state-of-the-art intrusion detection systems, achieving a ~30% higher success rate than previous methods.
Deep learning (DL)-based Network Intrusion Detection System (NIDS) has demonstrated great promise in detecting malicious network traffic. However, they face significant security risks due to their vulnerability to adversarial examples (AEs). Most existing adversarial attacks maliciously perturb data to maximize misclassification errors. Among AEs, natural adversarial examples (NAEs) are particularly difficult to detect because they closely resemble real data, making them challenging for both humans and machine learning models to distinguish from legitimate inputs. Creating NAEs is crucial for testing and strengthening NIDS defenses. This paper proposes NetDiffuser1, a novel framework for generating NAEs capable of deceiving NIDS. NetDiffuser consists of two novel components. First, a new feature categorization algorithm is designed to identify relatively independent features in network traffic. Perturbing these features minimizes changes while preserving network flow validity. The second component is a novel application of diffusion models to inject semantically consistent perturbations for generating NAEs. NetDiffuser performance was extensively evaluated using three benchmark NIDS datasets across various model architectures and state-of-the-art adversarial detectors. Our experimental results show that NetDiffuser achieves up to a 29.93% higher attack success rate and reduces AE detection performance by at least 0.267 (in some cases up to 0.534) in the Area under the Receiver Operating Characteristic Curve (AUC-ROC) score compared to the baseline attacks.
