Software-defined wide area networks (SD-WAN), empowered by software-defined networking (SDN) technology, offer unparalleled flexibility and efficiency in wireless communication. However, their integration introduces new security challenges, particularly in mitigating distributed denial of service (DDoS) attacks. In this paper, we propose an advanced security framework tailored to SDN-enabled SD-WAN. A dataset collected during experiments was used for training and testing the model’s performance. Our framework leverages machine learning algorithms to detect and classify DDoS attacks targeting SD-WAN controllers, considering the environment’s unique characteristics. We develop adaptive machine learning model capable of accurately discerning high-rate and low-rate DDoS attacks, enhancing the network’s resilience against sophisticated threats. Results from controlled experiments show promise for real-world deployment, though further validation is needed. Our results highlight the framework’s ability to adapt to dynamic network conditions and provide robust security for SDN-enabled SD-WAN. Our adaptive model integrates RF and DT explicitly for SD-WAN contexts, achieving 99. 97% accuracy for high-rate attacks and 99. 96% for low-rate attacks. Our QT-PCA preprocessing pipeline reduces dimensionality while preserving performance, and PACKETIN event-triggered mitigation enables dynamic response. This method significantly enhances security solutions’ accuracy, ensuring robust DDoS attack detection in SD-WAN environments. Additionally, by leveraging the SD-WAN architecture’s efficiency, our approach optimizes network performance, underscoring its efficacy and practicality in enhancing security and efficiency.
Musa et al. (Tue,) studied this question.