Open Access

ARTICLE

DDoS Defense Model on 5G Network Slices

Kun-Lin Tsai¹, Shih-Ting Chiu², Chihhsiong Shih², Fang-Yie Leu^2,*

1 Electrical Engineering Department, Tunghai University, Taichung, Taiwan
2 Computer Science Department, Tunghai University, Taichung, Taiwan

* Corresponding Author: Fang-Yie Leu. Email:

(This article belongs to the Special Issue: Advanced Security and Privacy for Future Mobile Internet and Convergence Applications: A Computer Modeling Approach)

Computer Modeling in Engineering & Sciences 2026, 148(2), 46 https://doi.org/10.32604/cmes.2026.083958

Received 14 April 2026; Accepted 17 June 2026; Issue published 28 August 2026

Abstract

With the quick development of 5G networks, network slicing and Open Radio Access Network (O-RAN) have become key technologies for improving network resource-allocation efficiency and flexibility. However, network slicing also faces intrusion-detection challenges, particularly for detecting DDoS attacks, which are difficult to detect due to traffic being silently transmitted across multiple sub-slices. To address this problem, this paper proposes a 5G network slicing intrusion detection mechanism, called the DDoS Defense Model on 5G Network Slices (2D5NS) which integrates machine learning and real-time traffic monitoring techniques to detect and mitigate DDoS attacks within an O-RAN. This security system consists of a Random Forest (RF) classification model, which is deployed within the Service Management and Orchestration (SMO) of O-RAN to classify packets transmitted from UE to the RAN into eMBB, mMTC and uRLLC slices, and a detection approach comprising the XGBoost mechanism which monitors the traffic within each slice in real time to detect DDoS attacks issued by User Equipment (UE). Once traffic is abnormal, it triggers an Entropy Algorithm to identify the sources of the DDoS attacks. The simulation results of our second experiment show that the classification accuracies of RF classification model in its 3-fold Cross Validation (CV) for eMBB and mMTC training achieve 99.98%. In our third experiment, the detection accuracy of 2D5NS/XGBoost model on uRLLC reaches at least 93.43%. Several state-of-the-art systems are evaluated. Here, the conclusion is that the 2D5NS outperforms each of them and the 2D5NS can effectively mitigate and block DDoS attacks for network slices.

---

Keywords

---