sc 23(1): e4

Research Article

An Intelligent Machine Learning based Intrusion Detection System (IDS) for Smart cities networks

Download220 downloads
  • @ARTICLE{10.4108/eetsc.v7i1.2825,
        author={Muhammad Yaseen Ayub and Usman Haider and Ali Haider and Muhammad Tehmasib Ali Tashfeen and Hina Shoukat and Abdul Basit},
        title={An Intelligent Machine Learning based Intrusion Detection System (IDS) for Smart cities networks},
        journal={EAI Endorsed Transactions on Smart Cities},
        volume={7},
        number={1},
        publisher={EAI},
        journal_a={SC},
        year={2023},
        month={3},
        keywords={IoT, IDS, Machine learning},
        doi={10.4108/eetsc.v7i1.2825}
    }
    
  • Muhammad Yaseen Ayub
    Usman Haider
    Ali Haider
    Muhammad Tehmasib Ali Tashfeen
    Hina Shoukat
    Abdul Basit
    Year: 2023
    An Intelligent Machine Learning based Intrusion Detection System (IDS) for Smart cities networks
    SC
    EAI
    DOI: 10.4108/eetsc.v7i1.2825
Muhammad Yaseen Ayub1,*, Usman Haider2, Ali Haider1, Muhammad Tehmasib Ali Tashfeen3, Hina Shoukat1, Abdul Basit1
  • 1: COMSATS University Islamabad
  • 2: National University of Computer and Emerging Sciences
  • 3: FAST National University, Peshawar, Pakistan
*Contact email: mryaseenayub@gmail.com

Abstract

INTRODUCTION: Internet of Things (IoT) along with Cloud based systems are opening a new domain of development. They have several applications from smart homes, Smart farming, Smart cities, smart grid etc. Due to IoT sensors operating in such close proximity to humans and critical infrastructure, there arises privacy and security issues. Securing an IoT network is very essential and is a hot research topic. Different types of Intrusion Detection Systems (IDS) have been developed to detect and prevent an unauthorized intrusion into the network. OBJECTIVES: The paper presents a Machine Learning based light, fast and reliable Intrusion Detection System (IDS). METHODS: Multiple Supervised machine learning algorithms are applied and their results are compared. Algorithms applied include Linear Discriminant analysis, Quadratic Discriminant Analysis, XG Boost, KNN and Decision Tree. RESULTS: Simulation results showed that KNN Algorithm gives us the highest accuracy, followed by XG Boost and Decision Tree which are not far behind. CONCLUSION: A fast, secure and intelligent IDS is developed using machine learning algorithms. The resulting IDS can be used in various types of networks especially in IoT based networks.