Multi-Target Tracking based on the EK-GLMB Filter in 3D Using Active Radar with Velocity Measurement

The measurement collected by 3D (3-dimensional) active radar is usually a non-linear polar coordinate measurement and capture true targets and false clutter information. Furthermore, the true targets are usually unknown and time-varying and the correlation between targets and measurements is uncertain. Hence, multi-target tracking in 3D is extremely challenging. In this paper, we firstly establish 3D polar coordinate measurement model. Specifically, for the radar which can measure line-of-sight (LOS) rate and target Doppler information, we build 3D velocity-measurement model by modifying the former. Then, combined with the non-linear measurement model, a tracking algorithm based on the extended Kalman (EK) generalized labeled multi-Bernoulli (GLMB) filter is proposed for multitarget tracking in 3D. Finally, numerical results demonstrate the validity of the proposed method in 3D using active radar in the presence of detection uncertainty, false observations, and noise.