Skip to content
SDGs Center UDINUS

Evidence record · UDINUS · SDGs

Optimizing Intrusion Detection Performance in the Internet of Vehicles through De-duplication and SMOTEENN on an Imbalanced Dataset

Rektorat UDINUS 2025 Published 2026-10-30

Evidence record · UDINUS · SDGs

Unit
Rektorat UDINUS
Period
2025
Goals
Record
7ed625c6-ed01-4dc0-83d7-3218f92c8acb
Published
2026-10-30

Universitas Dian Nuswantoro conducted research on improving the performance of intrusion detection systems in the Internet of Vehicles, with Hakim, Lukmanul; Ghozi, Wildanil; Sani, Ramadhan Rakhmat; Yassin, Warusia; Handoko, L. Budi; and Rafrastara, Fauzi Adi participating in the activity. The research examined the Internet of Vehicles as a focused implementation of Internet of Things technologies in the automotive domain. In this environment, vehicles communicate or exchange data in real time with surrounding components to improve mobility efficiency and road safety. The study also addressed the susceptibility of the Internet of Vehicles to cyber intrusions, notably Spoofing and Denial-of-Service attacks.

The research proposed data-processing methods to improve the performance of intrusion detection systems using the Controller Area Network communication protocol. The CICIoV2024 dataset was utilized because it captures realistic vehicular traffic under benign and malicious scenarios. Preprocessing involved eliminating duplicate records and handling class imbalance with the SMOTEENN technique to reduce model bias. Six machine learning models were evaluated: XGBoost, Random Forest, Logistic Regression, AdaBoost, Deep Neural Networks, and Support Vector Machines. Accuracy, precision, recall, and F1-score were used as performance metrics.

Experimental findings showed that the proposed approach significantly improved detection accuracy. Random Forest delivered the highest performance, followed by XGBoost. The activity therefore addressed the development of intelligent, adaptive, and secure intrusion detection frameworks for Internet of Vehicles environments. It relates to the stated goal concerning connected and intelligent transportation through its focus on vehicular communication, road safety, and cybersecurity. The activity presents Universitas Dian Nuswantoro research on data processing and machine learning for intrusion detection in connected transportation environments.

Reference https://www.scopus.com/pages/publications/105036001826?origin=resultslist