CFAR detection based on PDF estimation of clutter

In this paper we propose a new CFAR (constant false alarm rate) detector whose detection threshold is calculated based on the PDF (probability density function) estimation of clutter. Two schemes for the proposed detector are introduced, which use the Padé approximation method and the maximum entropy (MaxEnt) method respectively to estimate the PDF of clutter. The two schemes are verified in Weibull clutter through simulation experiments, and the results show that, the proposed CFAR detector performs as well as the maximum likelihood (ML) CFAR detector in uniform Weibull clutter, and it can adapt to other types of clutter without changing the detector´s structure.