The week's pick
Reseach Article
Towards Real-Time DoS Detection: A Multi-Objective Optimized SVM Framework using Kernel Approximation and Dimensionality Reduction
| International Journal of Computer Applications |
| Foundation of Computer Science (FCS), NY, USA |
| Volume 187 - Number 102 |
| Year of Publication: 2026 |
| Authors: Loubna Ali, George Nartey Debrah, Youssef Ali |
10.5120/ijca7b271ca4ab8b
|
Loubna Ali, George Nartey Debrah, Youssef Ali . Towards Real-Time DoS Detection: A Multi-Objective Optimized SVM Framework using Kernel Approximation and Dimensionality Reduction. International Journal of Computer Applications. 187, 102 ( May 2026), 7-14. DOI=10.5120/ijca7b271ca4ab8b
Abstract
Denial-of-Service (DoS) attacks remain one of the most critical threats to modern network infrastructures, requiring intrusion detection systems (IDS) that are both highly accurate and computationally efficient. While Support Vector Machines (SVM) have demonstrated strong performance in detecting cyber attacks, their high computational complexity and long training time limit their applicability in real-time environments. This paper proposes a unified lightweight framework for real-time DoS detection based on a hybrid optimization of SVM. The framework integrates Principal Component Analysis (PCA) for dimensionality reduction, the Nystr¨om method for kernel approximation, and a linear SVM classifier to achieve nonlinear decision boundaries with significantly reduced computational cost. A multiobjective Bayesian optimization strategy is employed to jointly optimize key parameters, including feature dimension, kernel approximation size, and SVM hyperparameters, with the objective of maximizing detection recall while minimizing training time and model complexity. The proposed framework is evaluated on three benchmark intrusion detection datasets: UNSW-NB15, CIC-IDS2017, and BoTIoT, representing diverse network environments and attack distributions. Experimental results demonstrate that the optimized framework consistently improves detection performance while significantly reducing computational cost. Notably, the model achieves up to 99.97% recall on the BoT-IoT dataset while reducing training time by over 97%. On CIC-IDS2017, recall improved from 0.9331 to 0.9868, representing an absolute increase of 5.37 percentage points, while training time was reduced by 96%. These results confirm that the proposed approach effectively balances detection accuracy and computational efficiency, making it highly suitable for real-time intrusion detection systems. Furthermore, the consistent performance across multiple datasets demonstrates the generalizability and robustness of the proposed framework.
References
- L. Ali, C. B. Njima, D. Balaganesh, W. Alhasan, and A. Ali, “Cybersecurity – Enhancing digital transaction authentication through improved digital signature process,” in 2025 International Conference on Control, Automation and Diagnosis (ICCAD), Barcelona, Spain, 2025, pp. 1–6. doi: 10.1109/ICCAD64771.2025.11099180.
- L. Ali, S. Hajnulla, and N. Souliman, “Reducing the wireless sensor networks delay by reducing program complexity and using parallel processing mechanisms,” EMSJ Journal, 2022.
- L. Ali, H. Mathieu, and F. Biennier, “Monitoring and managing distributed networks using mobile agents,” in Proceedings of the 2nd International Conference on Information & Communication Technologies (ICTTA), Damascus, Syria, 2006, pp. 3377–3382. doi: 10.1109/ICTTA.2006.1684959.
- H. Mathieu, L. Ali, and F. Biennier, “A distributed management system: Towards proactive information system management in virtual enterprises,” in 12th IFAC Symposium on Information Control Problems in Manufacturing, Saint Etienne, France, 2006, pp. 659–665. Available: https://hal.science/hal-00196078.
- F. Biennier, L. Ali, and A. Legait, “Extended service integration: Towards manufacturing SLA,” in Advances in Production Management Systems, IFIP, vol. 246, Springer, Boston, MA, 2007. doi: 10.1007/978-0-387-74157-4 11.
- C. Cortes and V. Vapnik, “Support-vector networks,” Machine Learning, vol. 20, no. 3, pp. 273–297, 1995. doi: 10.1007/BF00994018.
- N. Cristianini and J. Shawe-Taylor, An Introduction to Support Vector Machines. Cambridge: Cambridge University Press, 2000. doi: 10.1017/CBO9780511801389.
- W. Wang, M. Zhu, X. Zeng, X. Ye, and Y. Sheng, “Malware traffic classification using convolutional neural network for representation learning,” in IEEE International Conference on Information Networking (ICOIN), 2017. doi: 10.1109/ICOIN.2017.7899468.
- B. Sch¨olkopf and A. J. Smola, Learning with Kernels. MIT Press, 2002. Available: https://mitpress.mit.edu/9780262194754/learning-withkernels/.
- L. Ali and F. Biennier, “Integration of security requirements in virtual enterprises,” in APMS Conference, 2005. Available: https://hal.science/hal-00393896/.
- I. T. Jolliffe, Principal Component Analysis. Springer, 2002. doi: 10.1007/b98835.
- C. K. I. Williams and M. Seeger, “Using the Nystr¨om method to speed up kernel machines,” in Advances in Neural Information Processing Systems, 2001. Available: https://papers.nips.cc/paper/2000/hash/ 19de10adbaa1b2ee13f77f679fa1483a-Abstract.html.
- A. Rahimi and B. Recht, “Random features for large-scale kernel machines,” in Advances in Neural Information Processing Systems, 2007. Available: https://papers.nips.cc/paper/2007/hash/ 013a006f03dbc5392effeb8f18fda755-Abstract.html.
- M. Ring, D. Wunderlich, D. Scheuring, D. Landes, and A. Hotho, “A survey of network-based intrusion detection data sets,” Computers & Security, vol. 86, pp. 147–167, 2019. doi: 10.1016/j.cose.2019.06.005.
- L. Ali, M. Jaber, S. Chaari, and F. Biennier, “Context-aware infrastructure to support distributed industrial services,” in Proceedings of the IEEE International Conference on Emerging Technologies and Factory Automation (ETFA), 2007, pp. 716–719.
- N. Issa, D. Gruska, and L. Ali, “A hybrid GAM-based model for predicting vulnerability exploitation,” in Cooperative Information Systems (CoopIS 2025), Lecture Notes in Computer Science, vol. 15535, Springer, Cham, 2026. doi: 10.1007/978-3-032-15538-2 29.
- J. Snoek, H. Larochelle, and R. P. Adams, “Practical Bayesian optimization of machine learning algorithms,” in Advances in Neural Information Processing Systems, 2012. Available: https://papers.nips.cc/paper/2012/hash/ 05311655a15b75fab86956663e1819cd-Abstract.html.
- L. Ali, Gestionnaire d’infrastructure distribu´ee. Ph.D. dissertation, Institut National des Sciences Appliqu´ees de Lyon, 2008.
- N. Moustafa and J. Slay, “UNSW-NB15: A comprehensive data set for network intrusion detection systems,” in Military Communications and Information Systems Conference (MilCIS), 2015. doi: 10.1109/MilCIS.2015.7348942.
- I. Sharafaldin, A. H. Lashkari, and A. A. Ghorbani, “Toward generating a new intrusion detection dataset and intrusion traffic characterization,” in Proceedings of ICISSP, 2018. doi: 10.5220/0006639801080116.
- N. Moustafa, “The BoT-IoT dataset,” Data in Brief, vol. 24, p. 103386, 2019. doi: 10.1016/j.dib.2019.103386
Index Terms
Keywords