CFAR Detector in Clutter Edge Situation Using Stationary Wavelet Transform

In this paper, a new wavelet-based constant false alarm rate (CFAR) detector called wavelet-CFAR (W-CFAR), composed of the clutter edge detector and the cell-averaging (CA)-CFAR processor, is proposed in the clutter edge situation. The proposed detector does not require any prior knowledge about the background environment. It uses wavelet transform for edge detection and then selects appropriate part of reference cells for thresholding. Although CA-CFAR is an optimal detector for the homogeneous environment, but its performance is degraded in the non-homogeneous environment. The performance of the proposed detector is evaluated and compared with those of the Greatest Of (GO), Smallest Of (SO), and CA-CFAR detectors. The simulation results show that the proposed detector provides low loss CFAR performance for the homogenous environment in comparison with CA-CFAR. The clutter edge may be stayed on before or after cell under test (CUT). The performance of W-CFAR approaches to SO-CFAR as long as the position of clutter edge is before CUT; otherwise its performance approaches to GO-CFAR.