An exoatmospheric chaff cloud tracking method based on the GMM-VB-EKF strategy
Y. K. Shen, Bin Rao, Weijie Wang, Boyu Han, Shaowei Su, Yekui Qian, Hao Huang · IET conference proceedings. · 2026
This paper presents a novel Gaussian Mixture Model-Variational Bayesian-Extended Kalman Filter (GMM-VB-EKF) approach for exoatmospheric chaff cloud tracking in ballistic missile defense. Conventional multi-target tracking methods show limitations when handling dense, dynamic passive radar interferences like chaff clouds. The proposed method employs an extended target tracking framework incorporating: a velocity-dependent chaff motion model for elliptical diffusion characterization, an adaptive DBSCAN-GMM clustering algorithm addressing new target detection constraints in variational Bayesian methods, and a hybrid VB-EKF algorithm for spherical coordinate state prediction. Simulation experiments suggest this approach effectively track dynamic chaff clouds while performing target counting and morphological feature extraction.