Autonomous Flight Path Planning for Traffic Monitoring in Wireless Sensor Networks

Abstract

In this research, ground-based wireless sensor networks and unmanned aircraft are used to determine optimal flight paths to ensure longer dwell times over areas of high activity. Sensor nodes placed in rugged, austere, or dangerous locations may be difficult to access. Ground-based data collection agents can be limited by range and terrain. Multi-hopping data between sensor nodes are subject to communication range limits, while energy reserves of nodes handling the highest data traffic decay at a faster rate. The unmanned aircraft is ideally suited to overcome these limitations. The scenario proposed in this research utilizes the unmanned aircraft to fly an optimal flight path within the wireless sensor network, collect their data, determine which sensor nodes are experiencing the most activity in their sensing range. Once the sensor nodes with the highest activity rates are identified, a new optimal flight path is calculated to ensure longer flight times times over those particular sensors, while maintaining or decreasing flight times over sensors with less or no activity. When multiple unmanned aircraft are employed, sensors are clustered by euclidean distances to a centroid by a k-means clustering algorithm. Optimal flight paths are determined by solving an optimal control problem using direct collocation methods. Single and multiple unmanned aircraft solutions are presented.