Maneuvering Target Parameter Estimation Based on Sparse BayesianDictionary Learning in Space-Time Adaptive Processing br

A sparse Bayesian dictionary learning-based parameter estimation method is proposed to overcomethe performance degradation in presence of dictionary mismatch in Space-Time Adaptive Processing (STAP).First, multiple measurements are constructed by using direction compensated space samples. Second, thebilinear transformation is utilized to separate the velocity and acceleration of the maneuvering target. Finally,the dynamic dictionaries of velocity and acceleration are established by the Taylor's series, and then themaneuvering target parameters are estimated by sparse Bayesian dictionary learning. Numerical results showthat the proposed method can obtain better accuracy in parameter estimation, and can provide an improvedperformance to the sparse recovery methods with pre-discretized dictionary in STAP parameter estimation