Cluster-Based Random Forest Regression for Internet of Things Bandwidth Prediction Using Leakage-Free Features

Naek Parulian Hutagalung; Bambang Tutuko; Ahmad Zarkasi

doi:10.57152/malcom.v6i3.2894

Authors

Naek Parulian Hutagalung Universitas Sriwijaya
Bambang Tutuko Universitas Sriwijaya
Ahmad Zarkasi Universitas Sriwijaya

DOI:

https://doi.org/10.57152/malcom.v6i3.2894

Keywords:

Bandwidth Prediction, Feature Importance, IoT Traffic, K-Means Clustering, Random Forest

Abstract

Heterogeneous Internet of Things (IoT) deployments generate highly variable network traffic, yet most existing bandwidth management approaches rely on static provisioning or device-agnostic models that fail to capture per-device traffic heterogeneity. This study proposes a hybrid machine learning framework combining K-Means clustering and Random Forest regression to enable cluster-aware, leakage-free bandwidth consumption prediction for IoT environments. Using 994,145 flow records from the UNSW HomeNet dataset, the Log1p transformation was applied to address heavy-tailed traffic distributions, and direct bandwidth-derived features were strictly excluded to prevent data leakage. Devices were segmented into three traffic profiles (Low, Medium, and High), validated by a Silhouette Score of 0.4153 and a Davies–Bouldin Index of 1.0357. Within the structural constraints of this single-dataset, controlled setting, Random Forest regression achieved R² values above 0.99 across all clusters, with stable cross-validated performance, substantially outperforming Linear Regression, which achieved R² values as low as 0.0374 in the Medium cluster. Feature importance analysis revealed that Inter-Arrival Time is the primary driver of bandwidth consumption, superseding raw payload volume, with direct implications for temporally-driven feature design. These results indicate that cluster-aware, leakage-free models offer a viable architecture for IoT bandwidth management, with cross-domain generalization as a primary direction for future work.

Downloads

Download data is not yet available.

References

S. H. Alsamhi et al., “Machine Learning for Smart Environments in B5G Networks: Connectivity and QoS,” Comput. Intell. Neurosci., vol. 2021, no. 1, p. 6805151, Jan. 2021, doi: https://doi.org/10.1155/2021/6805151.

M. M. Rahman, F. Bouhafs, S. A. Hoseini, and F. den Hartog, “UNSW HomeNet: A network traffic flow dataset for AI-based smart home device classification,” Comput. Ind. Eng., vol. 204, no. C, Art. no. 111041, Jun. 2025, doi: 10.1016/j.cie.2025.111041.

Z. Ali et al., “Edge Based Priority-Aware Dynamic Resource Allocation for Internet of Things Networks,” Entropy, vol. 24, no. 11, Art. no. 1607, 2022, doi: 10.3390/e24111607.

O. Aouedi, V. A. Le, K. Piamrat, and Y. Ji, “Deep Learning on Network Traffic Prediction: Recent Advances, Analysis, and Future Directions,” ACM Comput. Surv., vol. 57, no. 6, Art. no. 151, pp. 1–37, Feb. 2025, doi: 10.1145/3703447.

I. Lohrasbinasab, A. Shahraki, A. Taherkordi, and A. Delia Jurcut, “From statistical- to machine learning-based network traffic prediction,” Transactions on Emerging Telecommunications Technologies, vol. 33, no. 4, p. e4394, Apr. 2022, doi: https://doi.org/10.1002/ett.4394.

M. T. Abdelaziz et al., “Enhancing Network Threat Detection with Random Forest-Based NIDS and Permutation Feature Importance,” Journal of Network and Systems Management, vol. 33, no. 1, p. 2, 2024, doi: 10.1007/s10922-024-09874-0.

N. Derakhshanfard, L. Hosseinzadeh, F. R. Jafari, and A. Ghaffari, “Intelligent resource allocation in internet of things using random forest and clustering techniques,” Sci. Rep., vol. 15, no. 1, p. 30522, 2025, doi: 10.1038/s41598-025-15931-8.

K. Saurabh, S. Mahapatra, and M. M. Tripathi, “An adaptive bandwidth scheduling algorithm for IoT devices using machine learning techniques,” Franklin Open, vol. 14, Art. no. 100492, 2026, doi: https://doi.org/10.1016/j.fraope.2026.100492.

C. Pan, Y. Wang, H. Shi, J. Shi, and R. Cai, “Network Traffic Prediction Incorporating Prior Knowledge for an Intelligent Network,” Sensors, vol. 22, no. 7, Art. no. 2674, 2022, doi: 10.3390/s22072674.

K. Murphy, A. Lavignotte, and C. Lepers, “Fault Prediction for Heterogeneous Telecommunication Networks Using Machine Learning: A Survey,” IEEE Transactions on Network and Service Management, vol. 21, no. 2, pp. 2515–2538, 2024, doi: 10.1109/TNSM.2023.3340351.

W. Symbor and ?. Falas, “Ensuring Reliable Network Communication and Data Processing in Internet of Things Systems with Prediction-Based Resource Allocation,” Sensors, vol. 25, no. 1, Art. no. 247, 2025, doi: 10.3390/s25010247.

F. X. K. Koesnadi and S.-H. Chung, “Improved Cell Allocation Strategies Using K-Means Clustering in Congested 6TiSCH Environments,” Sensors, vol. 24, no. 17, Art. no. 5608, 2024, doi: 10.3390/s24175608.

Y. Xiong, S. Dong, R. Liu, F. Shi, and X. Jing, “IoT network traffic classification: a deep learning method with Fourier transform-assisted hyperparameter optimization,” Front. Phys., vol. 11, Art. no. 1273862, Oct. 2023, doi: 10.3389/fphy.2023.1273862.

F. Afifi, F. Zaki, H. Hanif, N. Aqil, and N. B. Anuar, “Transformer-based tokenization for IoT traffic classification across diverse network environments,” PeerJ Comput. Sci., vol. 11, p. e3126, Aug. 2025, doi: 10.7717/peerj-cs.3126.

V. A. Phan, J. Jerabek, and L. Malina, “Comparison of Multiple Feature Selection Techniques for Machine Learning-Based Detection of IoT Attacks,” in Proc. 19th Int. Conf. Availability, Reliability and Security (ARES '24), Vienna, Austria, Jul. 2024, Art. no. 157, pp. 1–10, doi: 10.1145/3664476.3670440.

M. A. Bouke and A. Abdullah, “An empirical study of pattern leakage impact during data preprocessing on machine learning-based intrusion detection models reliability,” Expert Syst. Appl., vol. 230, Art. no. 120715, Nov. 2023, doi: https://doi.org/10.1016/j.eswa.2023.120715.

A. Apicella, F. Isgrò, and R. Prevete, “Don’t push the button! Exploring data leakage risks in machine learning and transfer learning,” Artif. Intell. Rev., vol. 58, no. 11, p. 339, 2025, doi: 10.1007/s10462-025-11326-3.

B. Bischl et al., “Hyperparameter optimization: Foundations, algorithms, best practices, and open challenges,” WIREs Data Mining and Knowledge Discovery, vol. 13, no. 2, p. e1484, Mar. 2023, doi: https://doi.org/10.1002/widm.1484.

S. Pundir, G. Bhandari, and S. Tyagi, “Dynamic Bandwidth Allocation Using Linear Regression Model in IOT with Machine Learning Techniques,” in Micro-Electronics and Telecommunication Engineering, D. K. Sharma, S. L. Peng, R. Sharma, and G. Jeon, Eds., Lecture Notes in Networks and Systems, vol. 617. Singapore: Springer, 2023, pp. 597–609. doi: 10.1007/978-981-19-9512-5_54.

J. Cohen, Statistical Power Analysis for the Behavioral Sciences, 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988.

R. Kulandaivel, R. MANIKANDAN, S. Veerappampalayam Easwaramoorthy, and J. Cho, “Gauss Markov and Flow Balanced Vector Radial Learning network traffic classification on IoT with SDN,” PLoS One, vol. 19, no. 10, Art. no. e0308052, Oct. 2024, doi: 10.1371/journal.pone.0308052.