DTXG-RF-based Intrusion Detection System for Artificial IoT Cyber Attacks
Shayma Wail Nourildean, Wafa Mefteh, Ali Frihida · Engineering Technology & Applied Science Research · 2025
The swift advancement of networking technology and the rising incidence of cyber-attacks have made effective cybersecurity a critical priority. The primary concern with IoT networks is their susceptibility to vulnerabilities. IoT security necessitates the substantial involvement of artificial intelligence as a security technology to mitigate these challenges. Cyberattacks are evolving in sophistication, consequently posing greater obstacles in the precise detection of intrusions. An Intrusion Detection System (IDS) is a device or software application that monitors the activities of network systems for malicious actions or policy breaches and produces reports. The primary objective of an IDS is to efficiently identify attacks. Moreover, it is imperative to identify attacks at an early stage to mitigate their effects. Machine learning models have become increasingly popular in IDSs due to their capacity to process substantial data volumes and identify patterns in real time. Machine learning involves building an algorithm to identify consistent patterns within a dataset. This study aimed to build an IDS using an ensemble machine learning (DTXG-RF) model and compare it with DT, XGBoost, KNN, RF, NB, and CatBoost on the CIC-IoT-2023 and a Ransomware dataset. The results showed that the proposed DTXG-RF outperformed other machine learning models with accuracy reaching 95.06%.