A CFAR Detector Based on Elastic Net for HFSWR

In practical scenarios, target detection for high-frequency surface wave radar (HFSWR) is challenging because it is performed in a complex multi-target detection background. This paper proposes a constant false alarm rate (CFAR) detector based on the elastic net to address the issue of target detection in multi-target backgrounds for HFSWR. The objective function of the proposed CFAR detector not only considers the goodness of fit of the clutter model to the clutter but also incorporates a regularization term. Specifically, the Weibull distribution is used to fit the HFSWR clutter. Additionally, due to the sparsity of the targets, an elastic net-based regularization term is incorporated, which is flexible, stable, and effectively captures sparsity. By minimizing the objective function, estimates of the target echo in each cell can be obtained. Cells containing targets are then excluded based on these estimates, while the remaining cells are used as reference cells to estimate the background distribution parameters. The proposed CFAR detector performs well in both simulations and experiments on real data, demonstrating its effectiveness and robustness.