Robust Radar Tracking via Joint Adaptive EKF and Power Allocation Under Dynamic Clutter

ABSTRACT In this letter, a joint adaptive extended Kalman filter (AEKF) and power allocation strategy is proposed for robust multitarget tracking in colocated MIMO radar systems under dynamic clutter. The unknown clutter intensity is incorporated into the measurement model and characterized by a truncated inverse gamma distribution. Using variational Bayesian inference, the proposed AEKF jointly estimates the target state and clutter intensity. Based on these estimates, a power allocation strategy is developed using the predicted conditional Cramér–Rao lower bound as a performance metric. Simulation results demonstrate that the proposed method achieves improved tracking performance in dynamic clutter environments.