基于双剔除门限的Switching-CFAR目标检测算法

为了解决传统的目标检测算法在非均匀噪声环境下检测性能严重下降的问题，提出了一种基于双剔除门限的Switching-CFAR（switching-constant false alarm rate based on dual censoring thresholds，DCS-CFAR）目标检测算法。基于参考窗参考单元样本期望值和测试单元，得到双重剔除功率比较门限。通过双重比较，剔除参考窗中极大值参考单元，根据剩余参考单元数，选择合适的参考单元来估计背景噪声功率，并得到功率检测门限。在Matlab环境下，通过蒙特卡洛方法和SwerlingII模型对DCS-CFAR目标检测算法的关键参数，以及在各种仿真环境下与其它目标检测算法的检测性能进行了仿真对比分析，DCS-CFAR目标检测算法在均匀背景噪声下，检测率为98.76%，接近于CA-CFAR算法；在杂波和多干扰目标环境下，检测率分别为97.83%和98.23%。在均匀和非均匀噪声环境下，DCS-CFAR检测算法均优于ACCA-CFAR和GO-CFAR算法。结果表明，提出的DCS-CFAR检测算法在均匀和非均匀噪声环境下, 均具有良好的检测性能。

Switching-CFAR target detection algorithm based on dual censoring thresholds

In order to solve the problem that the detection performance of the conventional CFAR detector degrades severely in non homogenous environment, a switching-constant false alarm rate target detection algorithm(DCS-CFAR) based on dual censoring thresholds is proposed. The dual censoring thresholds are obtained based on the cell under test and the mean power level of the reference window. Compared with the dual censoring thresholds, the higher amplitude reference cells are censored from the reference window. By comparing with the eliminate threshold, reference maximum cell in the reference window is removed. Based on the number of the remaining cells in the reference window, the appropriate reference cells are chosen to estimate the background noise power level. The optimal detection threshold is obtained through multiplying estimated value of background noise power level by the factor that is calculated based on the target false alarm rate. For the Swerling II targets model in homogeneous and non-homogenous noise environments such as multiple interferences and clutter edges, Monte-Carlo simulations are presented to obtain the key parameters and demonstrate the detection performance of the DCS-CFAR and other target detection algorithms using Matlab software. Through the simulation comparison with other target detection algorithms, detection rate of DCS-CFAR detection algorithm is 98.76% in homogeneous environment and close to the CA-CFAR algorithm. Its detection rates are 97.83% and 98.23% in non-homogeneous environments, respectively. The DCS-CFAR detection algorithm is superior to ACCA-CFAR and GO-CFAR algorithms in homogeneous and non-homogeneous environments. The results show that the proposed DCS-CFAR detection algorithm not only provides low CFAR loss in homogenous environment but also performs robustly in non-homogenous environments.