Energy-Efficient UAV Trajectory and Data Offloading Optimization for LoRa-Enabled IoT Networks

Recent advancements in Unmanned Aerial Vehicles (UAVs) have led to their widespread adoption across diverse operational domains. Their extended flight endurance and increasing onboard computational capabilities make them particularly valuable in communication systems, especially in urban, industrial, and poorly connected environments. In parallel, the Internet of Things (IoT) has emerged as a complementary technology, enabling sensing and data offloading through ad hoc protocols such as LoRaWAN. The effective integration of UAVs and IoT, however, raises new challenges in trajectory planning and energy-efficient data transmission, particularly under the strict operational constraints of LoRaWAN networks. To address these challenges, we introduce the Path-and-Offloading Problem (POP), a new combinatorial optimization problem that jointly considers UAV trajectory design and data offloading. POP is formulated both as a mixed-integer nonlinear programming (MINLP) model and as a mixed-integer linear programming (MILP) variant. The objective is to minimize the total energy consumption of UAV operations and sensor transmissions, while ensuring compliance with LoRaWAN transmission policies and completing data offloading from sensors to a single UAV. Our formulation, which incorporates realistic energy and communication models, optimizes the UAV trajectory, its hovering locations and durations, and key transmission parameters of LoRa-enabled sensors. Computational experiments across diverse scenarios demonstrate the effectiveness and scalability of the proposed models, providing practical insights into UAV-enabled IoT networks.