False alarm control of CFAR algorithms with experimental bistatic radar data

Abstract

False alarm control performance of different constant false alarm rate (CFAR) algorithms is experimentally investigated using bistatic radar data. The CFARs under investigation include cell-averaging (CA-CFAR), smaller-of (SO-CFAR), greater-of (GO-CFAR), ordered-statistic (OS-CFAR), censored cell-averaging (CCA-CFAR), and the homogeneity detector mean-to-mean ratio (MMR) test. The experimental data was collected using an Illuminator of Opportunity (IOO) bistatic radar that is under ongoing development at DSTO. For this set of experimental data, it is observed that CFAR algorithms that have a higher CFAR loss in a homogeneous environment will have a larger theory-experiment false alarm mismatch, especially at smaller false alarm rates and larger CFAR window sizes.