A Hybrid SDN Architecture for IDS Using Bio-Inspired Optimization Techniques

Abstract

Software-defined networking (SDN) is a networking paradigm of subsequent generation where various network components are used by a centralized controller that allows reliability in network system configuration, execution of policy decisions, and management via a primary programmable network infrastructure unit. SDN is known to deny DDoS attacks despite the default security protocols. State-of-the-art researches have shown that SDN intrusion is possible in diverse layers of its generalized architecture. Addressing this problem, this work presents an optimized intrusion detection system for SDN to mitigate the effect of DDoS attacks. This article’s main contribution comprises the development of a voting strategy-based ensemble classifier, which is established based on bio-inspired particle swarm optimization and salp swarm optimization in the context of optimized classification of DDoS attack-prone traffic SDN. Experimental analysis of the proposed SDN-IDS depicts that the proposed strategy outperforms existing classifiers in terms of accuracy.