A Flying Ad-Hoc Networks Framework for Smart Farming and Precision Agriculture

This paper presents a novel deployment and data routing framework for precision agriculture using a Flying Ad Hoc Network (FANET) architecture that integrates both aerial and ground-based sensing nodes. To ensure optimal environmental coverage with minimal sensing radius, a centroidal Voronoi partitioning of the agricultural field is performed, where Chebyshev centers serve as the generators of each partition cell. This approach guarantees uniform sensing regions, balanced load distribution, collision-free aerial operations, and reduced transmission power requirements. Ground sensing nodes within each cell collect essential soil parameters such as moisture, temperature, pH, electrical conductivity, macronutrient levels (N, P, K), and organic matter content. UAVs gather ambient weather data and act as mobile relays, forwarding both atmospheric and ground-level observations to quasi-stationary sink nodes located at high altitudes. These sink nodes are equipped to capture vertical environmental profiles and function as edge computing platforms that perform real-time localized predictions on agricultural indicators, such as irrigation demand and crop health status. Furthermore, using an opportunistic routing paradigm—including flooding, probabilistic forwarding, and contact-based heuristics—both localized predictions and sensor data are reliably transferred to the main base station. This enables centralized global forecasting and application-layer analytics. The proposed system thus supports scalable, energy-efficient, and adaptive smart farming solutions under realistic mobility and energy constraints.