CFAR Detection For Wide-band Coherent Rader In Nonhomogenous Sea Clutter

For anti-ship missiles, primary task of radar seeker is implementing detection ofsea targets correctly and efficiently. The harsh natural environment, complex interferingenvironment as well as multi-targets environment seriously restricted detectionperformance of sea surface targets for the radar seeker. How to improve the ability ofguided weapons to adapt to the complex battlefield environment has become a hot issuein current domestic and international research.Based on broadband coherent radar system, this dissertation takes attacking typicallarge ship targets as application background, and carries out the theory and technique ofvalid estimation of clutter parameters under nonhomogeneous background, parametricconstant false alarm rate (CFAR) detection under interfering or multi-targetsenvironment and non-parametric CFAR detection in nonhomogeneous background. Themain works of this dissertation can be summarized as follows:In the introduction, background and significance of the research are set forth fromthe view of requirements firstly, and then the research status are summarized from threeparts: characteristics of broadband target and clutter, target CFAR detection in clutterbackground, broadband target detection, at last, it gives brief introduction to thecontents and structural arrangements of the study.Chapter II accomplishes echo simulation for scenes of detection above sea surface.First, Multiple scattering centers model are adopted for target, fluctuation characteristicsof radar cross section (RCS) for targets are analyzed and the applicability of Swerlingmodels is discussed, still as well as slow moving targets echo models are achieved.Secondly, we introduce backscattering coefficient, amplitude statistical properties andrelated characteristics of sea clutter in this part. Finally, the echo is generated forfollow-up work to provide simulation data.In chapter III, we have a study of the adaptive parametric target CFAR detectionproblem under interfering and multi-targets environment. Firstly, causes fornonhomogeneous background are analyzed and the uniformity hypothesis testing isintroduced. Subsequently, we explore the automatically censoring algorithm for clutterreference cells based on ordered data variability, discuss the threshold and analyze theperformance of excluding interference. Finally, an adaptive target detection algorithm isproposed and simulation results show the validity and effectiveness of the algorithm.The chapter IV studies the non-parametric CFAR detection technology undercomplicated nonhomogeneous background. We first analyze the effect of signalprocessing on probability density function (PDF) of raw sea clutter, and discuss themodel mismatch problem in parametric CFAR. Then we introduced two methods to estimate clutter PDF, i.e. padé approximation and the maximum entropy method. Atlast, non-parametric CFAR detector is proposed, and simulation analysis areimplemented with compound K clutter; moreover, contrast with traditional parametricCFAR detectors demonstrates the performance superiority and robustness of thedetector under unknown background.