An Effective and Efficient Long-time Coherent Integration Method for Highly Maneuvering Radar Target in Sparse Domain

Robust and effective detection of "high-speed and highly maneuvering" target is one of the fundamental and difficult problems in both military and civil fields. Long-time coherent integration (LTCI) is proved to be an effective way to strengthen the weak signal and improve signal-to-clutter ratio. Recently, we have extended the concept of LTCI and proposed several methods for maneuvering target detection and estimation, i.e., Radon-fractional Fourier transform (RFRFT), Radon-linear canonical transform (RLCT), Radon-fractional ambiguity function (RFRAF), Radon-linear canonical ambiguity function (RLCAF), and phase differentiation-Radon-Lv's distribution (PD-RLVD). They can compensate the range and Doppler migrations simultaneously while the computational burden is the biggest problem for real applications. In this paper, the concept of sparse time-frequency analysis is introduced. Then we tried to establish the concept of LTCI in sparse domain, which is called sparse LTCI (SLTCI). The SLTCI combines the merits of LTCI and compressive sensing (CS), which is effective for maneuvering target detection and efficient for computation. Finally, we give an example of marine target detection using CSIR data, which indicates that the proposed method can achieve higher integration gain, better clutter suppression ability, and less computational burden.