Drone Security for Precision Agriculture by Using One-Dimensional Convolutional Neural Network

Abstract

New advancements in agricultural techniques, methods of food production, and delivery have introduced new and relatively unexplored cyber-attack pathways, the security and economic implications of which are not yet fully understood. Precision agriculture is key to overcome predicted food supply shortages to fulfil global demand. A growing number of technologies, such as sensors, transmitters, and data systems, are used in smart farming environments to make decisions based on data. These decisions are then integrated with improved machinery to increase production and decrease input–output inconsistencies. Unmanned Aerial Vehicles (UAVs) are independent devices used in smart farming for various purposes. These devices are susceptible to different types of attacks. In this paper, we proposed a deep learning model for detecting attacks on UAVs by using a 1D Convolutional Neural Network. The NSL-KDD dataset is used to measure the performance of the proposed model, and remarkable accuracy of 99.77% and an impressively low false positive rate of 0.0038 is achieved.