Clutter Suppression for OTHR Based on Joint Low Rank and Group Sparse Representation

Zirui Chen,Yifei Ji,Yongsheng Zhang,Zhen Dong,Weijian Liu,Alei Chen,Junqiang Song

DOI: https://doi.org/10.1109/taes.2024.3452690

IF: 3.491

2024-01-01

IEEE Transactions on Aerospace and Electronic Systems

Abstract:Sea or ground clutter is one of the main bottleneck factors for target detection in skywave over-the-horizon radar (OTHR). Conventional methods can partly suppress clutter, however, they typically suffer from two shortcomings. Firstly, weak targets beside or within the clutter tend to be submerged, which increases the probability of miss-detection. Secondly, a large amount of isolated clutter may be residual, which increases the probability of false alarm (PFA). In this paper, a clutter suppression method based on joint low rank and group sparse representation (JLRGSR) is proposed to address these issues. The properties of OTHR echo signals are firstly analyzed, including the low rank of clutter in the range-slow time domain, the column sparsity of clutter in the range-Doppler (RD) domain, and the group sparsity of targets in the RD domain, based on which the objective function is constructed. It is observed that clutter is primarily characterized by the dual constraints of low rank and column sparsity. Furthermore, the group sparsity constraint helps maintain the target morphology's integrity and reduces isolated clutter residuals. The objective function is solved through the alternating direction method of multiplier (ADMM) to realize clutter suppression. Finaly, the effectiveness and superiority of the JLRGSR, that is, higher probability of detection and lower probability of false alarm, are experimentally verified by simulation data and real data