The Study Of Multi-Strategy Constant False Alarm Rate Detection Method Under Complex Clutter

As an important part of the radar receiver, a CFAR detector sets the thresholdadaptively by estimating the background clutter power. The traditional CFARdetectors only use a single CFAR strategy in the whole process with low flexibilityand additional CFAR loss. There are two main streams of approaches in the CFARliterature: the first stream focuses on the modifications of the traditional CFAR tomake a compromise on all conditions; the second stream aims to either combine theindividual algorithms proposed in the first stream in order to use their advantages incertain situations or detect the presence of non-homogeneity in the CFAR windowprior to applying suitable CFAR processing. This paper makes research into thesecond stream Multi-strategy CFAR algorithm based on the analysis of thebackground. The main work is displayed as following.Firstly, the present research situation and basic radar CFAR detection theory isanalyzed which is consisted of the summarize of present CFAR processing at homeand abroad, the model of the radar clutter and target for CFAR detection and thereview of conventional CFAR detectors under the background of Rayleighdistribution.Secondly, based on the VI-CFAR and S-CFAR, a switching variability index(SVI) CFAR detector is proposed under the background of Rayleigh distribution.The structure and parameters of SVI is presented. Theoretical analysis andsimulation results show that SVI maintains the good performance of VI inhomogeneous and clutter edge background, meanwhile, greatly improves thecapacity of anti-multi targets of S-CFAR.Finally, the Multi-strategy CFAR algorithm under non-Rayleigh is researched.Common clutter distributions and relevant parameter estimation methods aresummarized. Three common methods of distribution model identification, KS test,PDF test and Anderson-Darling test, are studied, and the results show the lastmethod has the best performance. Based on K-sample Anderson-Darling hypothesistest, a modified Anderson-Darling CFAR detector is proposed to collect thehomogeneous clutter blocks. The results show MAD can improve the capacity ofanti-multi targets within clutter blocks and non-homogeneous bestriding clutterblock.