3D AOA Target Tracking Using Multiple Distributed UAVs with Unknown Self-Localizations and Dynamic Anchors

This paper focuses on the angle-of-arrival (AOA) moving target tracking in the three-dimensional (3D) space using multiple unmanned aerial vehicles (UAVs). Due to the practical limitations of the UAV payload and work environment, the positions of some UAVs are not available. We develop a distributed UAV system to solve the target tracking problem with position-known and -unknown mixture UAVs. Specifically, two UAVs that install self-localization modules, such as a global positioning system (GPS) sensor, are required as dynamic anchors to obtain the absolute coordinates of the target and UAVs. A modified distributed extended Kalman filter (DEKF) is proposed with the consideration of communication distance restriction. In addition, a UAV path optimization algorithm based on the Nesterov accelerated gradient method is developed to improve estimation accuracy. The properties and effectiveness of the proposed method are discussed and validated by simulation examples. The impacts of communication constraints on the estimation performance are shown.