Title: Methods of anomaly detection for the prevention and detection of cyber attacks

Authors: N. Girubagari; T.N. Ravi

Addresses: Department of Computer Science, Thanthai Periyar Government Arts and Science College, Tiruchirappalli, Tamil Nadu, India; Affiliated to: Bharathidasan University, India ' Department of Computer Science, Thanthai Periyar Government Arts and Science College, Tiruchirappalli, Tamil Nadu, India; Affiliated to: Bharathidasan University, India

Abstract: The idea of the 'smart city' has developed in response to the issues brought on by the rapid growth in urbanisation and population. Smart cities are interconnected. IoT and big data analytics enable smart city efforts. IoT devices connected to the always-on IoT network for several functions make unauthorised entry easier. This issue involves the safety system's ability to identify prior unidentified attacks. System behavioural modelling and unattended or semi-supervised machine learning could solve this problem. Machine learning model training datasets affect security system efficacy. Cyber-physical objects' security restrictions make system data inaccessible. These datasets have been constructed several times, but their reliability and completeness are questionable. Cyber attacks affect data privacy and security in connected IoT contexts, including smart infrastructure, communication, e-governance, etc. Cybersecurity requires a machine learning-based intelligent detection system. IoT anomaly detection finds odd behaviour. Many researchers have studied anomaly detection methods to detect and thwart data exchange cyber attacks. Existing technologies detected some cyber attacks, but others required newer, more powerful approaches. This article discusses the pros and cons of different methods and highlights the obstacles and research gaps that prevent anomaly detection approaches from reaching their full potential.

Keywords: smart city; big data; internet of things; IoT; cyber attacks; attacks detection; anomaly detection; machine learning.

DOI: 10.1504/IJIEI.2023.136097

International Journal of Intelligent Engineering Informatics, 2023 Vol.11 No.4, pp.299 - 316

Received: 24 Mar 2023
Accepted: 28 Jul 2023
Published online: 16 Jan 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article