A Real-World Dataset “IDSIoT2024” for Machine Learning/Deep Learning Based Cyber Attack Detection System for IoT Architecture

The Internet of Things (IoT) is an emerging technology evaluating its inception in all domains like industries, home automation, healthcare, agriculture, etc. The major challenge in IoT is securing the IoT devices and information associated with the IoT devices from hackers or unauthorized persons. These cyber security attacks can be detected early for timely interventions and reduce major risks. Machine Learning (ML)/Deep Learning (DL) has become an influential technique for automatic Attack or Intrusion Detection Systems (IDS), putting forward numerous advantages over traditional techniques. A well-structured, real-time IoT security dataset must be used to develop IDS using ML/DL approaches. The article proposes a Real-World dataset IDSIoT2024 consisting of 16,230,955 records to help researchers advance solutions for tackling cyber security in IoT networks. The records presented in the dataset were captured using the Wireshark tool in a real-world IoT architecture. The data collection has taken two months from the middle of June 2023 to August 2023. The proposed dataset can be employed in the field of Artificial Intelligence (AI) to develop ML/DL approaches and detect cyber-attacks in IoT networks. A large amount of data needs to be trained to an ML algorithm to get precise detection of attacks. Therefore, the proposed dataset with a vast number of records with different attacks can accommodate the researchers for the development of Intrusion Detection Systems.