Dynamic Task Reallocation in UAV-Assisted LoRaWAN Networks for Smart Farming

Smart farming increasingly relies on wireless sensor networks for environmental monitoring; however, collecting data from distributed sensors remains challenging in large rural fields where fixed gateways are infeasible. This paper proposes CRAFT (Cooperative Relay-Aware Flight Tasking), a decentralized dynamic task reallocation framework for multi-UAV LoRaWAN-based data muling. CRAFT partitions the monitored area into relay-assisted zones and enables unmanned aerial vehicles (UAVs) to cooperatively and adaptively reassign sensing tasks during flight based on workload disparity and received signal strength indication (RSSI). A model-based performance evaluation, derived from analytical mobility and communication assumptions, indicates that CRAFT reduces mission completion time by approximately 23%, significantly improves energy balance across UAVs, and maintains high packet delivery reliability compared to static partitioning and greedy reassignment baselines. These results highlight the potential of lightweight cooperative control for scalable and energy-efficient UAV-assisted LoRaWAN networks in smart farming applications.