This work presents an adaptive IDS framework coupling a tabular transformer encoder with a class balanced experience replay buffer that replays benign traffic at every update to stabilize decision boundaries, and investigates the buffer with overt label flipping and stealthy backdoor poisoning attacks.
Abstract
Network intrusion detection systems (IDS) trained on fixed traffic snapshots decay silently after deployment as threat distributions shift. Fine-tuning models on new attacks triggers catastrophic forgetting, while retraining from scratch is computationally infeasible. Replay-based continual learning counters this, but existing methods unrealistically confine benign traffic to a single early task and ignore the replay buffer as a potential attack surface. To address this, we present an adaptive IDS framework coupling a tabular transformer encoder with a class balanced experience replay buffer that replays benign traffic at every update to stabilize decision boundaries. We introduce the class-instance incremental (CII) scenario where benign flows reappear alongside new attacks as a more faithful stress test, and probe the buffer with overt label flipping and stealthy backdoor poisoning attacks. On the CICIDS2017 benchmark, our framework achieved 0.9994 accuracy under the traditional class incremental setup and 0.9989 under CII, with negligible forgetting, drastically outperforming sequential fine-tuning (0.0052), EWC (0.0324), LwF (0.0699), and iCaRL (0.8770) baselines. While injecting benign traffic into every experience proves essential for preventing forgetting, the replay buffer introduces critical vulnerabilities. Label-flipping collapses the model entirely (0.0053 accuracy at a 1% budget), and the backdoor maintains 0.97 overall accuracy while driving the attack success rate on trigger flows to 95%, evading standard monitoring. Ultimately, while a modest replay budget recovers near-joint-training performance, ensuring buffer integrity emerges as a strict operational requirement.
Deep learning based network intrusion detection systems (IDS) can achieve strong traffic classification performance, but their resilience to adversarial manipulation remains a critical concern. This study evaluates the adversarial robustness of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models in a multiclass intrusion detection setting using the Train_Test_Network dataset with ten traffic classes. The models were trained on true sliding flow-window sequences under a unified preprocessing pipeline to support fair comparison. Adversarial robustness was first assessed under a white-box Fast Gradient Sign Method (FGSM) setting and then broadened through additional FGSM and Projected Gradient Descent (PGD) stress testing. SHapley Additive exPlanations (SHAP) were further used to analyse explanation instability under clean and adversarial conditions, and explanation-drift features were evaluated as a secondary adversarial detection signal. Under clean evaluation, both models achieved strong and nearly identical performance, with accuracies of 0.9614 for LSTM and 0.9615 for GRU and weighted F1-scores of 0.9597 and 0.9598, respectively. Under the main FGSM condition, performance declined substantially: the LSTM achieved adversarial accuracy of 0.6094 and weighted F1-score of 0.6290 with an evasion rate of 37.38%, while the GRU achieved adversarial accuracy of 0.5130 and weighted F1-score of 0.5690 with an evasion rate of 47.02%. The broader robustness sweep showed that iterative PGD exposed stronger fragility than FGSM alone. SHAP analysis indicated that adversarial perturbation altered both prediction outcomes and local explanation structure. A learned explanation-driven detector improved over the rule-based baseline, while larger-scale validation confirmed that explanation drift remained informative, though not perfectly separable, at broader scale. Overall, the results show that strong clean performance does not imply adversarial robustness, and that explanation drift provides a useful auxiliary signal for adversarial monitoring in recurrent IDS models.
Elijah M. Maseno, Yanxia Sun, Zenghui Wang· International Journal of Inf...· 0 citations
Experimental results demonstrate that the proposed model achieves high detection accuracy, strong discriminative capability, and low false alarm rates across both datasets, confirming its effectiveness and scalability for next-generation cybersecurity applications.
The findings confirm that the proposed IDSaaS framework provides an efficient, scalable, and adaptive solution for real-time cloud intrusion detection and significantly enhances the reliability and resilience of modern cloud and industrial cybersecurity infrastructures.
Unik B. Lokhande, Kavita Sonawane· Journal of Cloud Computing· 0 citations
Traditional signature-based and static machine-learning-based intrusion detection systems (IDSs) have limited ability to adapt to changing traffic characteristics, and dynamic cloud-based infrastructures are increasingly vulnerable to new and zero-day attacks. Given this drawback, this study aims to develop a hybrid deep reinforcement learning framework for proactive cloud-network intrusion detection, called ShieldDRLNet. It employs a convolutional neural network and a long short-term memory encoder to obtain a spatiotemporal traffic representation and uses a Double Deep Q-Network agent for adaptive sequential decision-making. It features a latency-aware reward function that optimizes the detection correctness, reduces false alarms, penalizes for missed attacks and optimizes the response time. Training stability is achieved by using experience replay and target network updates. Experiments conducted at CICIDS2017 demonstrate that the accuracy, precision, recall, and F1-score of ShieldDRLNet are 97.2%, 95.6%, 96.8%, and 96.2%, respectively, with an average end-to-end detection latency of 4.7 ms per traffic window. It achieves better performance than the classical machine-learning, deep-learning, Transformer and reinforcement-learning baselines, following a shared chronological evaluation protocol. The proposed components also contribute to and generalize the results of ablation, sensitivity and statistical analyses, as well as cross-dataset analyses with UNSW-NB15. The results provide an encouraging deployment-oriented solution to low latency and adaptive intrusion detection in dynamic cloud environments for ShieldDRLNet.
S. Venkatramulu, Anitha Patil, K. R. Pradeep et al.· Discover Computing· 0 citations
A Kitchenham-informed systematic literature review methodology, this review synthesizes 186 studies published between 2018 and 2026 and develops a perturbation-realism taxonomy, ranging from feature-level manipulation to executable packet-level attacks, that clarifies when reported success corresponds to deployable risk.