Proactive Intrusion Detection in SDN Infrastructures Harnessing Machine Learning Predictions
J. Herrera Cabral, Augusto Neto, Helber Wagner Da Silva · 2024
In the rapidly evolving 5G connectivity era, cyber-security is critical for protecting mission-critical applications and sensitive data. The need for delivering mobile "killer apps" introduces significant cyber threats and vulnerabilities. The programmable capabilities and centralized Software-Defined Networking (SDN) management enhance protection in 5G mobile networks, but yields potential risks by the centralized infrastructure control, thus jeopardizing security and reliability. Recent research put forward using Machine Learning (ML) technique to address these intrusion detection and mitigation mechanisms preemptively. This paper proposes the predictive SDN Intrusion Detection (pSID) system, which lies in ML to predict and neutralize cyber threats in a proative approach. Evaluated in a realistic testbed, pSID effectively identifies and mitigates threats before causing significant harm, highlighting the importance of proactive cybersecurity measures within SDN frameworks.