Design and Analysis of Residual Learning to Detect Attacks in Intrusion Detection System

Abstract

Software-defined networking (SDN) is a new approach to networking that offers flexible,
manageable, low-cost, and rapidly evolving networks. A deep learning anomaly detection
solution that relies heavily on flows is highly recommended due to the critical nature of flows
in SDN. In this paper, we were able to create an intrusion-based model for locating threats
within an SDN network. We develop an intrusion-based model to detect the possible attacks in
SDN network. The study uses residual network to study the behaviour of attacks in SDN
network and thereby detecting the attacks in network flow. The simulation is conducted to test
the efficacy of the residual deep learning model over various attacks. The validation shows an
increased classification accuracy by classifying the attacks than the existing methods.