An improved CFAR detector using wavelet shrinkage in multiple target environments

In constant false alarm rate (CFAR) detection, the noise level in the cell under test (CUT) is estimated by combining the outputs of reference cells. In nonhomogeneous environments, estimator should be immune to the presence of nonhomogeneities among the reference cells. Here, we propose a wavelet CFAR detector that can automatically censors nonhomogeneities of reference cells. This detector is a combination of a pre-processor for denoising of input signal and a conventional CA-CFAR detector. The detection performance of WT-CA-CFAR processor is compared with the CA- and CML-CFAR detectors for homogeneous and multitarget situations. The results show that the WT-CA detector acts like the CA-CFAR in a homogeneous background and performs robustly in presence of interfering targets.