A DATASET AT A GLANCE...

Internet of Things (IoT)–An interconnection of several entities, like sensors, controllers, actuators, communication modules, computing devices and similar other units is gradually transforming the worldly applications to be smarter and automated. Hence, it is not a surprise that IoT can even transform and assist agriculture in various ways, such as: (i) Monitoring crop health for surplus yield; (ii) Examining soil nutrients for plant growth and survival; (iii) Predicting changing weather conditions based on historical data; (iv) Planning irrigation or fertilizer treatment and many more. However, security is a major concern as with all the IoT applications. An insecure agricultural IoT might lead to increased productivity losses or financial losses and even threaten an agriculture-based nation’s economy. Thus, intrusion detection in smart agriculture becomes a notable research field. If you are a researcher intending to enhance the security of Agricultural Internet of Things (AG-IoT) systems, then avail the “Farm-Flow | AG-IoT Security: Intrusion Detection in Smart Agriculture Dataset” from the Zenodo repository (https://zenodo.org/records/10964648). Comprising three months of network traffic from a real-world Agricultural IoT (AG-IoT) environment, this dataset contains about 1,310,000 instances, which are grouped under seven network attack types and one attack-free usual traffic. The seven attack types in the dataset include, BotNet DDoS, ICMP Flood, Arp Spoofing, HTTP Flood, MQTT Flood, UDP Flood, TCP Flood and Port Scanning, enabling an intrusion detection accuracy of 92.67%.

Note: Researchers availing the dataset has a citation request as:

• Rafael Ferreira, Ivo Bispo, Carlos Rabadão, Leonel Santos, and Rogério Luís de C. Costa (2025). Farm-flow dataset: Intrusion detection in smart agriculture based on network flows, Computers and Electrical Engineering, Volume 121, 109892, DOI: 10.1016/j.compeleceng.2024.109892