Constant false alarm rate detection based on estimating statistical distribution of non‐homogeneous sea clutter in sky‐wave over‐the‐horizon radar

Abstract

In sky-wave over-the-horizon radar (OTHR), there are some difficulties to efficiently detect targets in the non-homogeneous sea clutter by using conventional constant false alarm rate detectors. The amplitude of the clutter is supposed to follow the Weibull distribution in OTHR. Moreover, the characteristic of the clutter with a continuous region distribution in the range-Doppler (RD) map suggests that there exist the spatial relationship between neighbouring cells. The authors propose a rigorous algorithm to incorporate these relationships and the regularisation of interferences into the estimation of the shape parameter and scale parameter for every cell under test. This algorithm is mainly based on the consideration that the improvement of detection performance could be carried out by properly utilising the spatial information between neighbouring cells in the non-homogeneous Weibull clutter. Then, the adaptive detection threshold is calculated by the estimated parameters. The validity and effectiveness of the proposed algorithm are verified by simulations and the experiments in the real RD maps collected from an experimental sky-wave OTHR system.