An improved adaptive CFAR detector for non-homogeneous environment

The conventional variability index constant false alarm rate (VI-CFAR) algorithm experiences significant degradation in detection performance when interference targets are located on both sides of the reference window. To mitigate this challenge, this paper proposes an enhanced VITM-CFAR algorithm which integrates the trimmed mean CFAR (TM-CFAR) method, evaluating the clutter environment of the cell under test (CUT) and adaptive deleting the front and rear reference half-windows. The method further categorizes the clutter edge environment into high and low-power areas. Depending on the partial deletion of reference units, the smallest-of CFAR (SO-CFAR) method is employed in multi-target environment and low-power region of the clutter edge, while the greatest-of CFAR (GO-CFAR) method is utilized in high-power zones. Monte Carlo simulation experiments are conducted to assess the effectiveness of the VITM-CFAR method. The results indicate that the proposed method has low CFAR loss under homogeneous environment, and is more robust under non-homogeneous environment, showcasing superior performanc against multi-target interference and clutter edge protection.