Experimentation with Local Intrusion Detection in IoT Networks Using Supervised Learning
Christiana Ioannou, Vasos Vassiliou
Abstract
In this paper we are experimenting with an intrusion detection system (IDS) for IoT. The IDS under consideration is employing a machine learning techniques for detecting novel at-tacks in the IoT network. We examine detection based on Support Vector Machines (SVM). The detection models were trained and evaluated for Selective Forward and Blackhole network routing layer attacks using IoT-testbed data and achieved up to 99.8% Accuracy rates and 100% Recall values.
Topics & Concepts
TestbedIntrusion detection systemComputer scienceSupport vector machineInternet of ThingsRouting (electronic design automation)Artificial intelligenceMachine learningLayer (electronics)Supervised learningComputer networkData miningArtificial neural networkComputer securityOrganic chemistryChemistryNetwork Security and Intrusion DetectionEnergy Efficient Wireless Sensor NetworksSecurity in Wireless Sensor Networks