Adaptive Intrusion Detection System using Transformer-Based Neural Networks and Continual Learning Approach with Adversarial Investigation

Detecting adversarial evasion in deep learning intrusion detection systems using explainable AI

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.

Multi-Class Network Intrusion Detection: Machine and Deep Learning Benchmark with Live Stream Deployment

From Signature to Attention: Transformer-Powered Intrusion Detection Systems for Cybersecurity

Intrusion detection systems (IDSs) play a vital role in safeguarding modern computer networks against increasingly sophisticated and high-volume cyber threats. Recent progress in artificial intelligence, especially deep learning, has allowed IDSs to go from static rule-based systems to adaptive and data-driven security solutions. But traditional machine learning- and convolution-based IDSs often have trouble finding long-range dependencies and temporal correlations in large-scale network traffic. This makes detection less accurate and increases the number of false alarms. This challenge becomes more pronounced in heterogeneous and evolving network environments. To address this, experiments are conducted on two widely used benchmark datasets: CIC-IDS2017 for binary intrusion detection and CICIDS2018 for multiclass attack classification. These datasets represent realistic network traffic with diverse attack categories and severe class imbalance. The proposed methodology employs a Transformer-based intrusion detection framework incorporating sequence windowing, positional encoding, and multi-head self-attention to learn contextual traffic representations. The primary contribution of this study lies in systematically integrating sliding temporal windowing, positional encoding, and multi-head self-attention into flow-level intrusion modeling, accompanied by empirical ablation analysis and statistical validation across two large-scale CIC benchmark datasets. Performance is evaluated using accuracy, precision, recall, F1-score, ROC-AUC, and false alarm rate. 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.

Adaptive intrusion detection system for cloud security using deep learning

Cloud computing environments are increasingly vulnerable to sophisticated cyber threats due to the rapid growth of cloud-native infrastructures, high-volume network traffic, and evolving attack strategies such as zero-day exploits, adversarial attacks, and polymorphic malware. Traditional intrusion detection systems often struggle to maintain high detection accuracy, scalability, adversarial robustness, and real-time processing capability in dynamic cloud environments. To address these challenges, this study proposes an adaptive Intrusion Detection System-as-a-Service (IDSaaS) framework using a Modified Attention-Gate based BiLSTM-GRU architecture integrated with GAN-assisted synthetic anomaly generation. The proposed framework combines Bidirectional Long Short-Term Memory networks for temporal dependency learning, Gated Recurrent Units for computational efficiency, a custom attention mechanism for dynamic feature prioritization, a modified recurring gate structure for optimized information retention, and a Revamping Input Sequence Weighing Structure (RISWS) loss function for improved anomaly classification performance. Experiments were conducted using the CICIDS-2018, UNSW-NB15, SCAPY-based, and SCADANet datasets within an AWS cloud deployment environment. The proposed model achieved strong intrusion detection performance with 97.91% multi-class detection accuracy, 98.67% precision, 97.93% recall, and 98.18% F1-score while maintaining low inference latency and high throughput suitable for real-time deployment. Additional evaluation on the SCADANet industrial cybersecurity dataset achieved 99.69% test accuracy, further validating the adaptability of the framework across heterogeneous cloud and industrial network environments. Throughput analysis demonstrated stable processing capability of approximately 48,000 packets per second and 12,500 flows per second under high traffic conditions. 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.

A hybrid deep reinforcement learning framework for proactive cloud network intrusion detection using spatiotemporal feature learning

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.

Adversarial Evasion in Machine-Learning-Based Network Intrusion Detection: A Systematic Review, Threat Modeling, and Research Roadmap

Machine learning-based Network Intrusion Detection Systems (ML-based NIDS) have become central to modern cyber defense because they can learn complex traffic patterns and detect evolving malicious behavior beyond predefined signatures. However, their reliance on statistical feature representations also exposes them to adversarial evasion, where malicious traffic is modified at inference time to induce misclassification while preserving its operational intent. In network environments, such attacks are shaped by protocol semantics, feature dependencies, traffic timing, and functionality-preservation constraints, making adversarial evasion in ML-based NIDS distinct from adversarial examples in more continuous domains. This review synthesizes 186 studies published between 2018 and 2026—a fast-growing but fragmented body of work—using a Kitchenham-informed systematic literature review methodology. It introduces a lifecycle-aware, NIDS-specific threat model and organizes the literature into three domains—adversarial generation, robustness evaluation, and defense—developing a hierarchical taxonomy within each and conducting a two-tier critical analysis that separates domain-specific limitations from cross-cutting methodological, empirical, and operational constraints. The synthesis shows that much reported evasion success reflects diagnostic feature-space sensitivity rather than executable, deployable adversarial risk, and that many defenses demonstrate component-level robustness rather than system-level resilience. The review also develops a perturbation-realism taxonomy, ranging from feature-level manipulation to executable packet-level attacks, that clarifies when reported success corresponds to deployable risk. Finally, it derives evidence-grounded lessons learned and a future research roadmap for improving realistic evaluation, reproducible reporting, and deployable robustness in adversarial ML-based NIDS.