Abstract
Sparse recovery (SR) based space-time adaptive processing (STAP) has attracted much attention due to its small requirement of snapshots in the estimation of the clutter plus noise covariance matrix (CNCM). However, most of the existing SR STAP methods suffer from the grid mismatch effect of the dictionary matrix. In this paper, a novel grid-less total variation minimization (TVM) based STAP approach is proposed, which avoids the discretization of the spatial-temporal profile and possible mismatch of the spatial-temporal gird. The optimization problem is firstly introduced to estimate the clutter subspace steering vector by minimizing the defined atomic norm based on the TVM. Then the optimization problem is reformulated via utilizing the property of the radar space-time steering vector and approximation of the Bessel function. Finally, with a solution obtained by the optimization problem, a projection method is presented to obtain an accurate estimation of the CNCM. The proposed STAP method can be applied for both the side-looking and non-side-looking case. Numerical results validate its effectiveness compared with the other SR STAP methods.
Highlights
Space-time adaptive processing (STAP) is widely used in the area of airborne warning radar to suppress clutter for detecting the slow moving targets [1]
Based on Reed, Mallett and Brennan (RMB) rules [2], the sample matrix inversion (SMI) method achieves a 3dB signal to interference plus noise ratio (SINR) loss compared to the optimal STAP processor, while the number of independent and identically distributed (i.i.d.) snapshots is at least twice the system degrees of freedom (DOFs)
The low rank property of the clutter subspace is depicted by the Brennan’s rule [1], and we construct the optimization problem based on the total variation minimization (TVM)
Summary
Space-time adaptive processing (STAP) is widely used in the area of airborne warning radar to suppress clutter for detecting the slow moving targets [1]. A STAP processor is a space-time filter which minimizes the clutter output power maintaining the response of the desired signal. Based on Reed, Mallett and Brennan (RMB) rules [2], the sample matrix inversion (SMI) method achieves a 3dB signal to interference plus noise ratio (SINR) loss compared to the optimal STAP processor, while the number of independent and identically distributed (i.i.d.) snapshots is at least twice the system degrees of freedom (DOFs). In order to reduce the sample support needed, many algorithms have been proposed in recent decades. It can be divided into the following categories: dimension reduction and rank reduction algorithms, The associate editor coordinating the review of this manuscript and approving it for publication was Prakasam Periasamy
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call