IoT Intrusion Detection System based on Machine Learning Algorithms using the UNSW-NB15 dataset - Abstract

Aug 3, 2026

Upload by Eliana

The evolution of communication systems with the rise of IoT has increased the frequency of attacks, making connected-object security an emerging area that still lacks comprehensive preventive measures. At the network layer, Intrusion Detection Systems (IDS) analyze network data to identify abnormal behavior. This work implements multiple machine learning models to build an IDS using the UNSW NB15 dataset, including data cleaning, feature engineering, and comparison of classifiers to select the most effective one, with random forest performing well for rare anomalies.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Abstract

Introduction

Literature Survey

IoT Intrusion Detection System based on Machine Learning Algorithms using the UNSW-NB15 dataset Nogbou Georges ANOH1, Tiémoman KONE2, Joel Christian ADEPO3 , Jean François M’MOH4, Michel BABRI5 Research Scholar 1-4, Professor 5 Department of Computer and Digital Science Virtual University ofCôte d’Ivoire Abidjan Côte d’Ivoire ABSTRACT The evolution of communications systems with the advent of IoT is leading to an increase in attacks against them. This is due to the fact that the security of connected objects in the IoT is an emerging area which still requires preventive solutions against various attacks. At the network security level, Intrusion Detection Systems (IDS) are used to analyze network data and detect abnormal behavior in the network. In this work, we implemented different machine learning models to build an intrusion detection system based on the UNSW NB15 dataset. To do this, we did data cleaning and feature engineering on the data in the pre-processing phase. Then we used various models such as logistic regression, support vector machine (SVM) classifier, decision tree, random forest, eXtreme Gradient Boosting (XGBoost) in order to predict attacks. Finally, an intrusion detection system is trained on various machine learning algorithms and we selected the most effective model. Experiments were carried out on the UNSW-NB15 dataset and subsequently we compared other machine learning algorithms, and this means that the random forest model on important parameters has a clear advantage in the detection of rare abnormal behaviors. Key Words: Intrusion detection system, Machine learning algorithms, Random forest, SVM, UNSW-NB15 dataset. 1. INTRODUCTION The evolution of communications systems with the advent ofIoT is leading to an increase in attacks against them. This is due to the fact that the security of connected objects in the IoT is an area still in development which requires prevention solutions against various attacks. At the network security level, Intrusion Detection Systems (IDS) are used to analyze network data and detect abnormal behavior in the network [1] . These IDSs aim to recognize different attacks by analyzing different data sources, mainly network traffic and system event logs. To implement this concept of intrusion detection, specific tools are necessary: IDS (Intrusion Detection System) . They will automatically collect data representative of system activity (server, application, system, network), analyze it and alert administrators when signs of attack are detected. The different types of intrusion detection system can be classified into two categories [1]:(i) The network IDS or NIDS (Network based IDS), (ii) The System IDS or HIDS (Host based IDS) . One of the biggest challenges in intrusion detection is ensuring that warnings are only caused by real attacks and that each attack is an escalation alert for administrators. A complete intrusion detection system consists of several parts, each of which has a specific and essential task in the detection process. We distinguish the following different blocks: (i) Data source from which we can check if an intrusion is taking place, (ii) Detection engine which will analyze the data received from previous sources to report events, (iii) Response to detection. The main use of an IDS is to warn of intrusion. To achieve this, they have several features at their disposal, such as triggering alarms in the management interface or sending e-mails to security engineers. IDS performance remains mixed, given the high number of false alarms. In order to find solutions, the construction of innovative intrusion detection systems becomes necessary and must have means of assessing them. In this work, we will implement an intrusion detection system based on artificial intelligence in IoT. The designed system takes into account different machine learning models and a UNSW NB15 dataset. The rest of this work is structured into four (4) secti

FAQs

Why are intrusion detection systems important in IoT networks?

IoT adoption increases attack volume, and IDSs analyze network traffic to detect abnormal behavior and alert administrators when real attacks are suspected.

What dataset and preprocessing steps are used to build the proposed IDS?

The system uses the UNSW NB15 dataset and performs data cleaning and feature engineering during the preprocessing phase.

Which machine learning models are evaluated for predicting attacks?

The work evaluates logistic regression, SVM classifier, decision tree, random forest, and XGBoost, then selects the best-performing model based on experiments.

Aug 3, 2026

Upload by Eliana