CFAR Detection in Nonhomogeneous Weibull Sea Clutter for Skywave OTHR

In this letter, a constant false alarm rate (CFAR) detector is proposed for skywave over-the-horizon radar (OTHR) under Weibull distribution. First, an objective function, composed of a data fidelity term and regularization terms, is formulated in the Bayesian framework. The low rank prior of the distribution parameter matrix, smoothness of the parameters along the Doppler/range dimension, and sparsity of the target in the range-Doppler (RD) matrix are used to regularize the objective function. Then, the distribution parameters of all cells under test (CUTs) in the RD matrix are estimated by optimizing the objective function, and the abnormal CUTs with targets are identified and separated. Finally, target detection is performed by calculating the detection threshold through the threshold coefficient and estimated distribution parameters. Experiments on OTHR real data demonstrate that the proposed detector has a good CFAR property and significantly improved the probability of detection, especially in the case of multitarget and clutter edge.