Machine Learning Based SDN-enabled Distributed Denial-of-Services Attacks Detection and Mitigation System for Internet of Things

Abstract

AbstractAdvancements of Internet of Things (IoT) enhance the application spectrum of smart networking and demand intelligent security measurements against cyber-attacks. Recent integration of Software Defined Networking (SDN) in IoT environments provides better network management by decoupling of control plane from forwarding plane. An advanced SDN based network management also utilize the machine learning models to classify IoT network traffic at OpenFlow Switches with the coordination of SDN controller. In this paper, we propose a novel SDN-enabled Distributed Denial-of-Services attacks Detection and Mitigation System (SDN-DMS) which utilize SDN enabled security mechanism for IoT devices with support of machine learning algorithms to develop Distributed Denial of Services (DDoS) detection and mitigation system. SDN-DMS integrates Floodlight and Pox SDN controllers to reconfigure the OpenFlow switches in order to mitigate the detected DDoS attacks by advanced Support Vector Machine (SVM) algorithms of Lagrangian Support Vector Machine (LSVM), Finite Newton Lagrangian Support Vector Machine (NLSVM), Smooth Support Vector Machine (SSVM) and Finite Newton Support Vector Machine (NSVM). SDN-DMS measures the network behaviour of IoT devices by collecting the network traffic and classifying the traffic as normal and DDoS attacks by using an environment-specific dataset.KeywordsCyber-attacksInternet of ThingsDistributed denial of servicesSoftware defined networkingMachine learningDetectionMitigation