A privacy-preserving intrusion detection framework that operates exclusively on flow-level traffic metadata without deep packet inspection, which supports the practicality of explainable Transformer models for secure and scalable APT detection in encrypted 5G environments.
Abstract
The widespread adoption of end-to-end encryption in 5G networks limits the effectiveness of traditional intrusion detection systems that rely on payload inspection. This challenge is particularly critical for detecting Advanced Persistent Threats (APTs), which employ low-rate, long-duration, and stealthy communication strategies to evade conventional defenses. This study presents a privacy-preserving intrusion detection framework that operates exclusively on flow-level traffic metadata without deep packet inspection. Network packets are aggregated into bidirectional flows, from which temporal, statistical, and directional features are extracted to characterize behavioral patterns. A Transformer-based model with self-attention is employed to capture long-range dependencies across encrypted traffic sequences and identify subtle, temporally dispersed attack indicators. The framework is evaluated on a large-scale 5G-relevant dataset containing over one million flow records and compared against classical machine learning, ensemble, CNN, and LSTM models. Results demonstrate high recall and strong F1-score in distinguishing APT from benign traffic. Attention-based and feature-level explanations further reveal that prolonged communication, irregular timing gaps, and directional asymmetry significantly influence detection decisions. The findings support the practicality of explainable Transformer models for secure and scalable APT detection in encrypted 5G environments.
A deep learning-based intrusion detection framework that utilizes flow-level statistical features and a hybrid Convolutional Neural Network–Bidirectional Long Short-Term Memory (CNN–BiLSTM) architecture to capture both spatial and temporal characteristics of network traffic is proposed.
H. Hussein, Saad Raad, A. Qasim· Al-Noor Journal of Engineeri...· 0 citations
The growth of Internet of Things devices has expanded the number of attackable targets of advanced cyber threats by orders of magnitude, and the limited computational capabilities of IoT devices have made more traditional intrusion detection systems infeasible to execute at the edge. A Knowledge-Distilled Multi-Model Intrusion Detection and Prevention System is presented, which combines a Residual Graph Convolutional Network, Long Short-Term Memory classifier, and a Conditional Tabular Generative Adversarial Network enhanced with the student-only autoencoders through multi-loss knowledge distillation to create a high-performing Teacher ensemble and compresses its intelligence into an edge-deployed system. The system is evaluated using the IoT-23 benchmark of 325 million network flow records and attains 98.2% Teacher-Student fidelity, 0.94 Knowledge Retention Score, and 4.4x throughput increase with 82.7% reduction in the parameter and provides per-packet SHapely Additive exPlanation forensic explainability.
S. K, Menaka T K, N. R· 2026 4th International Confe...· 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.
An explainable deep learning framework evaluated across multiple heterogeneous cyber attack datasets, including Kitsune, Server-Based network data,enterprise logs, and Malware Traffic datasets, demonstrating the effectiveness of the proposed framework in handling heterogeneous network traffic while providing interpretable insights into model predictions.
A lightweight, explainable IDS that combines a 1D-CNN for spatial feature analysis with SHAP for model interpretation, yielding streamlined models that preserve over 93% F1-score and reduce computational overhead by more than 38%, facilitating millisecond-level inference on edge hardware.
Miracle Udurume, Vladimir V. Shakhov, Insoo Koo· Scientific Reports· 0 citations
HTTP Flood attacks remain difficult to detect because they operate at the application layer, resemble legitimate user requests, and generate burst-based temporal traffic patterns. Previous DDoS detection studies often rely on outdated datasets, process network flows as independent records, insufficiently address class imbalance, and provide limited interpretability for security analysts. This study proposes a sequence-aware and explainable deep learning framework for HTTP Flood detection using Long Short-Term Memory (LSTM). Reconstructed HTTP traffic from the UNSW-NB15 dataset was processed through proxy labeling, data cleaning, feature normalization, and sliding-window transformation to convert flow-level records into temporal sequences. Class weighting and SMOTE oversampling were evaluated to mitigate imbalance, while SHAP and LIME were used to explain model decisions. The proposed LSTM model achieved an attack recall of 94.8%, a false negative rate of 5.2%, balanced accuracy of 94.3%, MCC of 0.824, and ROC-AUC of 0.975. The results show that temporal representation improves detection of bursty HTTP Flood behavior, whereas class weighting provides a better balance between attack sensitivity and false-alarm control. Explainability analysis further confirms that the model relies on technically meaningful indicators, including packet rate, flow duration, traffic asymmetry, and service concentration. This framework supports interpretable early-warning detection for application-layer DDoS attacks.
