Development of a Contemporary Dataset for Intrusion Detection Systems in IoT-Integrated Network Topologies

ÖZKÖK T. B., ÖZKAN M., ASKERBEYLİ İ.

6th International Conference on Problems of Cybernetics and Informatics, PCI 2025, Baku, Azerbaycan, 25 - 28 Ağustos 2025, (Tam Metin Bildiri) identifier

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/pci66488.2025.11219794
  • Basıldığı Şehir: Baku
  • Basıldığı Ülke: Azerbaycan
  • Anahtar Kelimeler: Artificial Intelligence, Intrusion detection, IoT
  • Ankara Üniversitesi Adresli: Evet

Özet

The rapid proliferation of Internet of Things (IoT) deviceshas significantly increased the attack surface of modern networks. There are not many up-to-date intrusion detection datasets in the literature that include IoT traffic. In this study, a new dataset is proposed that is generated in a hybrid network topology that includes heterogeneous IoT nodes, traditional client-server architecture, and simulated attack scenarios. Using real-time traffic capture and modern attack emulation tools, a diverse dataset with labeled benign and malicious traffic types is compiled, including attacks such as Address Resolution Protocol (ARP) Poisoning, Denial of Service (DoS), Media Access Control (MAC) address flooding, Secure Shell (SSH) brute force attack, web scraping, scanning, and various exploitations. This dataset aims to contribute to next-generation intrusion detection systems (IDS), especially those that leverage machine learning.