DDoS Detection Based on PCA and Renyi Entropy to Secure SDN
Krishna Kanodia, Harsh Kumar, Sanjeev Patel · 2024
Software Defined Networks (SDNs) offer a flexible paradigm for network management but are susceptible to cyber threats, notably Distributed Denial of Service (DDoS) attacks. This paper introduces a novel methodology, PCA+Renyi, aimed at enhancing the detection performance of DDoS attacks in SDNs. The proposed PCA+Renyi model integrates Principal Component Analysis (PCA) with Renyi entropy, leveraging dimensionality reduction and entropy-based techniques to identify DDoS attacks in SDNs effectively. Subsequently, a comprehensive comparative study of DDoS detection strategies in SDNs is presented, focusing on the efficacy of various entropy measures and dimensionality reduction techniques. We assess the accuracy of various entropy measures, including Shannon, Tsallis, Renyi, and Interquartile Range (IQR), proposed model for identifying DDoS attacks. Our proposed PCA+Renyi model outperforms the existing techniques in terms of accuracy and precision on real-world network traffic datasets.