Optimizing UAV-Based Data Collection in IoT Networks With Dynamic Service Time and Buffer-Aware Trajectory Planning

Unmanned Aerial Vehicles (UAVs) have become vital tools for data collection in Internet of Things (IoT) networks, enabling efficient monitoring and information acquisition across various domains. However, UAV-assisted IoT networks often face significant challenges such as high data loss, latency, and resource inefficiency due to inadequate buffer management and dynamic service time (DST) allocation for IoT nodes. Existing approaches frequently overlook critical factors such as IoT nodes' energy levels, UAVs' energy constraints, and the dynamic data generation rates of IoT devices. To address these challenges, this article introduces a novel strategy for dynamically optimizing UAV trajectories by integrating real-time data from ground nodes and UAV energy levels. Given the DST and buffer constraints (BC) of IoT devices, optimizing UAV trajectories for data collection is a complex problem. We propose an optimization framework that strategically plans UAV trajectories to minimize service time at designated Rendezvous Points (RPs) and bypasses RPs when necessary to reduce the overall trajectory path, taking into account buffer status and dynamic service time adjustments. To efficiently solve this optimization problem, we employ a meta-heuristic technique known as Path Cheapest Arc with Guided Local Search (PCA-GLS) for UAV route planning within predefined time windows. Extensive simulations demonstrate the effectiveness of our proposed solution in optimizing UAV trajectories and improving data collection performance compared to existing algorithms such as BA-UAV, ACO-MS, and NSGA-II.