Deep Guard: A Novel Transformer-Based Framework for Real-Time Threat Detection in Heterogeneous Cyber Environments DOI Open Access

K R Pradeep,

B.N. Lakshmi,

M Varaprasad Rao

и другие.

International Journal of Computational and Experimental Science and Engineering, Год журнала: 2025, Номер 11(2)

Опубликована: Май 23, 2025

With evolving cyber threats in Internet of Things (IoT) and Industrial IoT (IIoT) networks, challenges with heterogeneous data dynamic attack patterns cannot be addressed using traditional intrusion detection systems (IDS). We present DeepGuard, a novel deep learning framework for these challenges. DeepGuard enhances space environments by utilizing transformer architecture augmented Adaptive Multi-Head Attention (AMHA), implements temporal encoding, anomaly-aware learning. propose an algorithm that varies attention mechanisms the event entropy level, which enables model to give more underlying while filtering out noise. Specifically, encoding allows express inter-event dependencies among samples practically, loss function based on makes sensitive uncommon patterns, leading its strong generalization capability unseen threats. implement TON_IoT dataset, where achieves 98.54% accuracy 98.88% AUC, outperforms existing models other three metrics, including accuracy, precision, recall. This shows model's robustness, generalizability, applicability work interface alone online large scale. It is suited deployment modern-day IIoT environments, considering complexity imbalanced nature data. In future, we plan optimize this edge devices federated privacy-preserving distributed training.

Язык: Английский

Deep Guard: A Novel Transformer-Based Framework for Real-Time Threat Detection in Heterogeneous Cyber Environments DOI Open Access

K R Pradeep,

B.N. Lakshmi,

M Varaprasad Rao

и другие.

International Journal of Computational and Experimental Science and Engineering, Год журнала: 2025, Номер 11(2)

Опубликована: Май 23, 2025

With evolving cyber threats in Internet of Things (IoT) and Industrial IoT (IIoT) networks, challenges with heterogeneous data dynamic attack patterns cannot be addressed using traditional intrusion detection systems (IDS). We present DeepGuard, a novel deep learning framework for these challenges. DeepGuard enhances space environments by utilizing transformer architecture augmented Adaptive Multi-Head Attention (AMHA), implements temporal encoding, anomaly-aware learning. propose an algorithm that varies attention mechanisms the event entropy level, which enables model to give more underlying while filtering out noise. Specifically, encoding allows express inter-event dependencies among samples practically, loss function based on makes sensitive uncommon patterns, leading its strong generalization capability unseen threats. implement TON_IoT dataset, where achieves 98.54% accuracy 98.88% AUC, outperforms existing models other three metrics, including accuracy, precision, recall. This shows model's robustness, generalizability, applicability work interface alone online large scale. It is suited deployment modern-day IIoT environments, considering complexity imbalanced nature data. In future, we plan optimize this edge devices federated privacy-preserving distributed training.

Язык: Английский

Процитировано

0