一种距离扩展目标的稳健检测算法

A Robust Detection Method for Range-spread Targets

针对距离扩展目标稀疏分布的一般情况,提出了一种改进的基于有序统计的广义似然比检测的稳健算法。该算法采用"或规则",对各种有效距离单元数量假设条件下得出的OS-GLRT检测算法的判决结果进行融合,从而克服了OS-GLRT检测算法对有效距离单元数量估计值比较敏感的问题。所得到的修正OS-GLRT检测算法无需估计有效距离单元数量。通过理论分析证明了恒虚警率性。通过Monte Carlo仿真进行性能分析,说明修正OS-GLRT检测算法的有效性和鲁棒性。由于修正OS-GLRT检测算法无需事先估计有效距离单元的数量,因此便于实际操作。

When the energy of the range-spread target return is not equable in all range cells,a robust detection method for range-spread target is proposed.The proposed detector is a modified generalized likelihood ratio test based on order statistics（OS-GL-RT）. The modified OS-GLRT adopts ＂ the rule of OR＂ and fuses the decision results of the OS-GLRT detectors with all the assumptions about the number of valid range cells to make a final decision.The resulting modified OS-GLRT doesn＇ t need apriority information a-bout the number of valid range cells,and can overcome the sensitivity of OS-GLRT detector to the number of valid range cells.It shows that the modified OS-GLRT is a constant false alarm rate（CFAR） detector.The performance assessment conducted by Monte Carlo simulation confirms the effectiveness and robustness of the proposed detector.As the modified OS-GLRT detector does not need to estimate the number of valid range cells,it is easy to work in practice.