Path Planning with Multiple UAVs Considering the Sensing Range and Improved K-Means Clustering in WSNs

Abstract

Recently, an Unmanned Aerial Vehicle (UAV)-based Wireless Sensor Network (WSN) for data collection was proposed. Multiple UAVs are more effective than a single UAV in wide WSNs. However, in this scenario, many factors must be considered, such as collision avoidance, the appropriate flight path, and the task time. Therefore, it is important to effectively divide the mission areas of the UAVs. In this paper, we propose an improved k-means clustering algorithm that effectively distributes sensors with various densities and fairly assigns mission areas to UAVs with comparable performance. The proposed algorithm distributes mission areas more effectively than conventional methods using cluster head selection and improved k-means clustering. In addition, a postprocessing procedure for reducing the path length during UAV path planning for each mission area is important. Thus, a waypoint refinement algorithm that considers the sensing ranges of the sensor node and the UAV is proposed to effectively improve the flight path of the UAV. The task completion time is determined by evaluating how the UAV collects data through communication with the cluster head node. The simulation results show that the mission area distribution by the improved k-means clustering algorithm and postprocessing by the waypoint refinement algorithm improve the performance and the UAV flight path during data collection.