International Journal of Image, Graphics and Signal Processing (IJIGSP)
IJIGSP Vol. 18, No. 5, 8 Oct. 2026
Cover page and Table of Contents: PDF (size: 1064KB)
Hybrid Metadata Fusion with Attention for Multiple IoT Device Classification using BiLSTM
PDF (1064KB), PP.76-92
Views: 0 Downloads: 0
Author(s)
B. Arathi
1,*
Ravindra Babu Kallam
1
Nagamani H. S.
2
Mathamsetti Kaivalya
3
K. Jamal
4
Boudhayan Bhattacharya
5
Index Terms
IoT Device Classification, Metadata Fusion, BiLSTM, Attention Mechanism, Deep Learning, Network Traffic Analysis, Temporal Modelling.
Abstract
Accurate IoT device classification is essential for secure and efficient network management in heterogeneous environments. However, existing approaches struggle with overlapping traffic patterns and limited generalization across dynamic device behaviours. This paper proposes a Metadata Fusion with Attention-based BiLSTM (MFA-BiLSTM) framework that integrates time-domain and statistical features using an adaptive attention mechanism. The model captures both sequential dependencies and distributional characteristics of network traffic, enhancing feature representation and classification robustness. Experiments conducted on the CIC-IoT-Dataset2022 demonstrate improved performance with an accuracy of 99.40%, precision of 99.39%, recall of 99.36%, and F1-score of 99.38%. The results indicate that the proposed approach achieves reliable and scalable IoT device classification while maintaining interpretability.
Cite This Paper
B. Arathi, Ravindra Babu Kallam, Nagamani H. S., Mathamsetti Kaivalya, K. Jamal, Boudhayan Bhattacharya, "Hybrid Metadata Fusion with Attention for Multiple IoT Device Classification using BiLSTM", International Journal of Image, Graphics and Signal Processing(IJIGSP), Vol.18, No.5, pp. 76-92, 2026. DOI:10.5815/ijigsp.2026.05.05
Reference
[1]H. El-Sofany., S. A. El-Seoud, O. H. Karam, et al. “Using machine learning algorithms to enhance IoT system security, Scientific Reports,” 2024, vol.14, pp.12077. doi: 10.1038/s41598-024-62861-y.
[2]M. H. Khan, R. C. Roy Chowdhury, M. S. H. T. Arman, R. A. Pulok, M. M. Hasan, A. J. Md Jakaria, and B. A. Hussain, “Deep learning based IoT device categorization using network traffic analysis, in Proc. 3rd Int. Conf.” Computing Advancements (ICCA ’24), 2025, pp. 423–428. doi: 10.1145/3723178.3723234.
[3]Q. Wan Din, W. I. S. and M. A Aamir, “generalized three-tier hybrid model for classifying unseen IoT Waseem, devices in smart home environments,” Scientific Reports, 2025, vol. 15, pp.35388. doi: 10.1038/s41598-025-19303-0.
[4]A. Aldhaheri, F. Alwahedi, M. A. Ferrag, and A. Battah, “Deep learning for cyber threat detection in IoT networks: A review,” Internet of Things and Cyber-Physical Systems, vol. 4, pp. 110–128, 2024. doi: 10.1016/j.iotcps.2023.09.003.
[5]J. Adjei, N. Z. Heywood, B. Nandy, and N. Seddigh, “Identifying IoT devices: A machine learning analysis using traffic flow metadata,” in 2024 IEEE Network Operations and Management Symposium (NOMS), 2024, pp. 1–7. doi: 10.1109/NOMS59830.2024.10575442.
[6]R. R. Chowdhury and P. E. Abas, “A survey on device fingerprinting approach for resource-constraint IoT devices: Comparative study and research challenges,” Internet of Things, vol. 20, 100632, 2022. doi: 10.1016/j.iot.2022.100632.
[7]N. Liao and J. Guan, “Multi-scale convolutional feature fusion network based on attention mechanism for IoT traffic classification,” International Journal of Computational Intelligence Systems, vol. 17, no. 1, 36, 2024. doi: 10.1007/s44196-024-00421-y.
[8]M. W. Hasan, “Design of an IoT model for forecasting energy consumption of residential buildings based on improved long short-term memory (LSTM),” Measurement: Energy, vol. 5, 100033, 2025. doi: 10.1016/j.meaene.2024.100033.
[9]P. J. Kumar and S. Neduncheliyan, “A novel optimized deep learning based intrusion detection framework for an IoT networks,” International Journal of Computational and Experimental Science and Engineering, vol. 10, no. 4, pp. 1169–1180, 2024. doi: 10.22399/ijcesen.597.
[10]P. Sinha, D . Sahu,. S. Prakash, et al., “A high performance hybrid LSTM-CNN secure architecture for IoT environments using deep learning, “Scientific Reports, 2025, vol. 15, pp. 9684. doi: 10.1038/s41598-025-94500-5.
[11]N. Karmous, Y. B. Dhiab, M. O. Aoueileyine, N. Youssef, R. Bouallegue, and A. Yazidi, “Deep learning approaches for protecting IoT devices in smart homes from MitM attacks,” Frontiers in Computer Science, vol. 6, 1477501, 2024. doi: 10.3389/fcomp.2024.1477501.
[12]G. Singal, M. Rawat, R. Kumar, and R. Kushwaha, “IoT network device classification using traffic log by applying machine learning techniques,” IEEE Internet of Things Journal, vol. 13, no. 7, pp. 12780–12788, 2026. doi: 10.1109/JIOT.2026.3652548.
[13]H. Tahaei, A. Liu, H. Forooghikian, M. Gheisari, F. Zaki, N. B. Anuar, Z. Fang, and L. Huang, “Machine learning for Internet of Things (IoT) device identification: A comparative study,” PeerJComput. Sci., 2025, vol. 11, pp. e2873. Doi: 10.7717/peerj-cs.2873.
[14]A. O. Salau and M. M. Beyene, “Software defined networking-based network traffic classification using machine learning techniques” Scientific Reports, 2024, vol.14, pp.20060.doi: 10.1038/s41598-024-70983-6.
[15]M. A. Aleisa, “Traffic classification in SDN-based IoT network using two-level fused network with self-adaptive manta ray foraging,” Scientific Reports, 2025, vol.15, pp.881.doi: 10.1038/s41598-024-84775-5.
[16]M.Alharby, “Evaluating machine learning approaches for multiple attack classification with improved computational efficiency in IoT networks,” Scientific Reports, 2025, vol.15, pp.39914. doi: 10.1038/s41598-025-23711-7.