DDoS Detection Framework Using Machine Learning with Adaptive AI and Moving Target Defence

DDoS attacks represent one of the major issues faced by current-day network systems. They are characterized by saturating servers with large amount of traffic and preventing normal services from being provided. For this reason, in the present research an innovative solution for identifying and mitigating the occurrence of DDoS attacks was developed. Such a solutional is characteristics by the application of intelligence and ML algorithms in order to analyze traffic flows in the network and identify their characteristics. By analyzing the traffic, the relevant parameters were established while machine learning algorithms such as established while machine learning algorithms such as Random Forest and XGBoost were used in order to differentiate normal traffic from the attackers one. The conclusions drawn from this study confirm the possibility of using ML algorithms in order to detect abnormal network activity associated with DDoS attacks. While some machine algorithms proved their efficiency in separating malicious and normal traffic, other such as Isolation Forest could distinguish abnormal traffic flow patterns regardless of the training process.  Combining these models increases the system’s capacity to identify DDoS attacks in dynamically evolving networks. The Adaptive AI-Based Defence Controller also contributes to increasing the efficiency of the proposed system, being able to respond to attacks from outside. In the case when an attack has been detected, the system is capable of performing different actions, such as monitoring suspicious request slowing down, analysing or blocking attackers IP addresses. In case of a serious attack, the system activates Moving Target Defence strategy, which causes changes in the network configuration making it more difficult for the attacker to continue their activity. In general, it is possible to say that this system represents a solution that allows detecting and mitigating DDoS attacks by means of implementing machine learning for attack detection and defence strategies.

[1] F. L. Becerra-Suarez, I. Fernández-Roman, and M. G. Forero, “Improvement of Distributed Denial of Service Attack Detection through Machine Learning and Data Processing,” Mathematics, vol. 12, no. 9, 2024, doi: 10.3390/math12091294. [2] N. S. Musa, N. M. Mirza, S. H. Rafique, A. M. Abdallah, and T. Murugan, “Machine Learning and Deep Learning Techniques for Distributed Denial of Service Anomaly Detection in Software Defined Networks - Current Research Solutions,” IEEE Access, vol. 12, no. January, pp. 17982–18011, 2024, doi: 10.1109/ACCESS.2024.3360868. [3] E. Altulaihan, M. A. Almaiah, and A. Aljughaiman, “Anomaly Detection IDS for Detecting DoS Attacks in IoT Networks Based on Machine Learning Algorithms,” Sensors, vol. 24, no. 2, 2024, doi: 10.3390/s24020713. [4] A. F. Al-Zubidi, A. K. Farhan, and S. M. Towfek, “Predicting DoS and DDoS attacks in network security scenarios using a hybrid deep learning model,” Journal of Intelligent Systems, vol. 33, no. 1, 2024, doi: 10.1515/jisys-2023-0195. [5] A. A. Alashhab et al., “Enhancing DDoS Attack Detection and Mitigation in SDN Using an Ensemble Online Machine Learning Model,” IEEE Access, vol. 12, no. April, pp. 51630–51649, 2024, doi: 10.1109/ACCESS.2024.3384398. [6] I. AlSaleh, A. Al-Samawi, and L. Nissirat, “Novel Machine Learning Approach for DDoS Cloud Detection: Bayesian-Based CNN and Data Fusion Enhancements,” Sensors, vol. 24, no. 5, 2024, doi: 10.3390/s24051418. [7] S. Ahmed et al., “Effective and Efficient DDoS Attack Detection Using Deep Learning Algorithm, Multi-Layer Perceptron,” Future Internet, vol. 15, no. 2, pp. 1–24, 2023, doi: 10.3390/fi15020076. [8] M. Mittal, K. Kumar, and S. Behal, “Deep learning approaches for detecting DDoS attacks: a systematic review,” Soft Computing, vol. 27, no. 18, pp. 13039–13075, 2023, doi: 10.1007/s00500-021-06608-1. [9] H. Elubeyd and D. Yiltas-Kaplan, “Hybrid Deep Learning Approach for Automatic DoS/DDoS Attacks Detection in Software-Defined etworks,” Applied Sciences (Switzerland), vol. 13, no. 6, 2023, doi: 10.3390/app13063828. [10] D. Kumar, R. K. Pateriya, R. K. Gupta, V. Dehalwar, and A. Sharma, “DDoS Detection using Deep Learning,” Procedia Computer Science, vol. 218, pp. 2420–2429, 2022, doi: 10.1016/j.procs.2023.01.217.