Enhancing Network Intrusion Recovery in SDN with machine learning: an innovative approach

Hammad, Mohamed; Hewahi, Nabil; Elmedany, Wael

doi:10.1080/25765299.2023.2261219

Cited by 6 publications

(1 citation statement)

References 38 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…a 99.99% detection accuracy rate, outperforming other ML models like LR, RF, SVM, and KNN. Likewise, Mohamed et al[74] introduced ML-based network intrusion recovery strategy, a novel technique for intrusion recovery in SDN that leverages traffic pattern analysis for strategic backup path selection, significantly reducing recovery time by up to 90% compared to traditional methods.The studies mentioned represent significant advancements in CPP and IDS within the SDN framework. Although these methods effectively enhance network efficiency, performance, scalability, and security, they are typically deployed in isolation.…”

mentioning

confidence: 99%

An Integrated Framework for Controllers Placement and Security in Software-Defined Networks Ecosystem

Sebopelo,

Isong

2024

J. Inf. Syst. Informatics

View full text Add to dashboard Cite

In the evolving landscape of Software-Defined Networking (SDN), the strategic placement of controllers poses a critical challenge that necessitate a precise balance between network performance and security. This paper presents an integrated framework for enhancing security and performance in SDN by combining controller placement and intrusion detection systems (IDS). Unlike existing solutions which were implemented disjointedly, we propose a holistic approach that leverages the proximity of controllers to network traffic for real-time threat detection, rapid response, and mitigation of security attacks. We employ an advanced clustering model for optimal controller placement, reducing costs and latency while ensuring reliability and balanced loads. In addition, we utilize k-nearest neighbour (KNN) for efficient anomaly detection in our IDS for improved network security. Experimental results confirm the framework’s effectiveness in strengthening SDN security and resilience. The enhanced-DBSCAN-based CPP model significantly minimized the cost, and latency, and ensured continuous operation in dynamic SDN environments while the KNN-based IDS shows effectiveness in improving threat detection capabilities, achieving high detection accuracy of 100% on the LAN dataset, outperforming other machine learning models such as Random Forest and Naïve Bayes. The indication is that strategic controller deployment, in conjunction with IDS, can significantly bolster threat detection, response times, and the overall security stance of the SDN environment.

show abstract

mentioning

confidence: 99%