A Novel Approach for Detection of DDoS Attacks in Software-Defined Networks Based on Grey Wolf Optimizer and Support Vector Machine

Abstract

Software-defined networks face attacks that hinder efficient network provision and prevent users from accessing systems. Attack detection is crucial for better service provision and system resilience. Existing SDN-based Distributed Denial of Service (DDoS) detection technologies suffer from low accuracy, which is attributed to inadequate feature extraction and resultsin elevated false negative rates. This study introduces a solution leveraging the Grey Wolf Optimizer algorithm for feature selection to enhance DDoS attack detection and categorization. Employing a novel binary Grey Wolf optimization and Support Vector Machine (SVM) classifier on the InSDN dataset for SDNs, the proposed approach demonstrates superior performance, achieving 100% accuracy and recall. Feature selection with Binary Grey Wolf yields a 97% F1-Score using the unimodal equation and 100% accuracy, 96% recall, and a 98% F1-Score with the multi-modal equation, underscoring its efficacy in bolstering SDN security against DDoS attacks.