Localization and tracking of RF emitting targets with multiple unmanned aerial vehicles in large scale environments with uncertain transmitter power

In this paper we study the localization and tracking of a radio frequency (RF) emitting target using multiple unmanned aerial vehicles (UAVs) over a large scale environment. Although localization of RF emitting targets using multiple measurements is a well studied problem, the standard approaches become inefficient when the signal power is uncertain and there is significant noise in the received signal strength (RSS) when the search environment is large scale. We present a localization and tracking architecture, where a data driven neural network model is used for estimating the unknown signal strength and extended Kalman filters are utilized for eliminating the RSS noise and increase the precision of target tracking performance. We present simulation results in a 10 × 10 km² search area, where 3 fixed wing UAVs localize and track a target with up to 28.3 m average error distance.