Enhanced Network Security: Real-Time Malicious Traffic Detection in SD-WAN Using LSTM-GRU Hybrid Model

In today's digital world, network administration has been revolutionized by Software-Defined Wide Area Network (SD-WAN) technology, which offers enhanced management, flexibility, and efficiency across distributed network topologies. This transition also increases susceptibility to security risks, including harmful traffic that could compromise network integrity. This paper proposes a method for real-time malicious traffic detection in SD-WAN networks employing a hybrid deep learning model that incorporates Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. The LSTM-GRU architecture is designed to retrieve both short- and long-term connections in network traffic data, hence improving the model's capacity to reliably identify unusual behavior that signifies security vulnerabilities. The model achieves an impressive detection accuracy of 98.83 %, which proves that it is effective at safeguarding SD-WAN environments. This suggested framework enhances SD-WAN network security by offering a dependable, real-time threat detection system that can be smoothly integrated into current network management protocols to reduce risks and maintain operational continuity.