JTF-BASED HIGH-RESOLUTION RADAR IMAGING OF MICROMOTION TARGETS FROM CORRUPTED DATA

Effective radar imaging based on the joint time-frequency (JTF) distribution has played significant roles in recognition, measurement, and cataloguing of micromotion targets in space. Sometimes, however, the radar returns are incomplete due to strong interference and occlusion, and the phase of the available samples is corrupted as a result of inaccurate motion compensation or atmospheric turbulence. To tackle this problem, this paper constructs the data-adaptive, nonparametric dictionary in the JTF domain. Then, it solves the optimal and sparse JTF distribution by iterative optimization using the theory of sparse signal representation and the least-square-error criterion. Particularly, this method applies the modified augmented Lagrangian method to reduce the computational complexity. The validity of the method has been proved by simulated data.