针对传统恒虚警(CFAR)算法在非均匀环境下,待检测单元(CUT)与参考窗的分辨单元不具有独立同分布(IID)特性,检测器性能出现剧烈下降的问题,提出一种新的CFAR检测器。该检测器首先引入一种M-N杂波边缘二元积累实现非均匀杂波边缘提取;然后,对数据平面内相邻杂波边缘内的数据,利用一种地形特征分类算法实现对地形的分类编号;最后,根据地形编号选择与CUT相同地形的分辨单元作为参考单元实现CFAR检测,则所选择的参考单元与CUT具有IID特性。利用实测数据验证M-N杂波边缘二元积累检测算法和地形特征分类算法的有效性。计算机仿真证明:文中提出的CFAR检测器的性能,比传统CFAR检测器的性能有明显提升。

As in heterogeneous environment the cell under test(CUT) do not have independent and identical distribution (IID) with the resolution cell of reference window, the performance of traditional constant false alarm (CFAR) algorithm degrades rapidly. To account for these issues, this paper proposes a new CFAR detector. Firstly the detector introduces an M-N clutter edge binary integration to obtain multiple clutter edges in heterogeneous environment. Then, a terrain classification algorithm is utilized to number the terrain of the data between the adjacent clutter edges in the data plane, ultimately based on the numbered terrains, the CFAR detector which can obtain resolution cells that contain the same terrain with CUT as the reference windows is proposed. The measured data is used to verify the proposed M-N clutter edge binary integration and terrain classification algorithms. Then the computer simulation is used to test the performance of the proposed CFAR detector, and the result reveals that compared to the traditional CFAR detector, the performance of the proposed CFAR detector has significant improvement in the heterogeneous environment.

TN957