RLFAT: A Transformer-Based Relay Link Forged Attack Detection Mechanism in SDN

SDN is a modern internet architecture that has transformed the traditional structure in recent years. By segregating control and data planes of network, facilitates centralized management, scalability, dynamism, programmability. However, this very feature makes controllers vulnerable to cyber attacks, which can cause network-wide crashes, unlike conventional networks. One most stealthy attacks face relay link forgery attack topology deception attacks. Such an result erroneous overall views for controllers, leading network functionality breakdowns even crashes. In article, we introduce Relay Link Forgery Attack detection model based on Transformer deep learning first time. The (RLFAT) detects by extracting features from flows received controllers. A dataset large number networks with different topologies was collected. Finally, Relay-based trained dataset, its performance evaluated using accuracy, recall, F1 score, AUC metrics. For better validation, comparative experiments were conducted some common models. experimental results show our proposed good detecting RLFA outperforms other