TITAN: Combining a bidirectional forwarding graph and GCN to detect saturation attack targeted at SDN

Area of Science:

• Computer Science
• Network Security

Background:

• Software-Defined Networking (SDN) offers programmability but introduces security vulnerabilities.
• Saturation attacks exploit SDN's architecture to overwhelm switch flow tables, disrupting network traffic.
• Existing graph-based detection methods suffer from reduced accuracy due to suboptimal graph generation.

Purpose of the Study:

• To propose TITAN, an effective bidirectional forwarding graph-based method for saturation attack detection in SDN.
• To address limitations in current graph generation techniques for attack detection.

Main Methods:

• TITAN defines flow forwarding rules and topology, designing flow statistical features.
• It generates bidirectional forwarding graphs using flow statistics (nodes) and traffic paths (edges).
• A Graph Convolutional Network (GCN) is employed to analyze these graphs for saturation attack detection.

Main Results:

• TITAN transforms network traffic flows into bidirectional forwarding graphs.
• The method demonstrates effective detection of saturation attacks in SDN environments.
• Experimental results show a detection accuracy exceeding 97%.

Conclusions:

• TITAN offers a robust and accurate solution for saturation attack detection in SDN.
• The bidirectional forwarding graph approach effectively captures relevant information for attack identification.
• This method enhances the security of SDN by mitigating saturation attack risks.