SADOE: Sequential-based angle-Doppler off-grid estimation with coprime sampling structures for space-time adaptive processing

Parameter estimation for space-time adaptive processing (STAP) usually encounters conflicts between estimation accuracy and algorithmic computation complexity in practical applications. In addition, due to the complication of mobile characteristics in STAP, it is often necessary to detect many prominent surface targets. Herein, we propose a sequential approach based on coprime sampling structures for off-grid angle-Doppler estimation in the framework of STAP, where both the accuracy and efficiency are well balanced. The estimation is accomplished via three sequential stages, where the initialized relative spatial-temporal map and closed-form solutions to spatial and temporal parameters based on the rotational invariant technique are subsequently formulated, eventually paired by a Euclidean distance minimization-based principle. Benefiting from such sequential implementation, the computational complexity can be greatly reduced, and the basis mismatch problem can be avoided in the meantime. Besides, coprime sampling structures enable expanded degrees of freedom in the case of fixed physical cost, adapting to more complicated actual scenarios. Simulation results demonstrate the effectiveness of the proposed algorithm from the aspects of both accuracy and efficiency along with the advance of DOFs.