Research On Radar-clutter-processing Technology Based On Wavelet Transform

In the VTS(Vessel Traffic System),the capability of radar clutter processing directly influences the performance of target-extracting and target-tracking of radar,so it has always been an important topic for researchers.Nowadays,parametric methods based on the statistical characteristic of the clutter,not only fail to detect the small targets effectively,but hardly cope with clutters of all sorts of statistical characteristics.Meanwhile,nonparametric methods can adaptively process the radar echoes,but have a bad pertinence for they do not make use of the transcendent information.Therefore,it has a profound theoretical meaning and practical value to study on a more efficient algorithm and implementation of processing of radar clutter.This thesis mainly applys the theory of wavlet analysis to the processing of radar clutter in order to work out the algorithm and implementing model,then realize it in FPGA.Firstly,using the conception of multi-resolution analysis,the radar echo is decomposed into the smoothing items representing the distributive characteristic of clutter energy space,and wavelet coefficient items representing the variance composition of clutter.Secondly,on the basis of the above theoretical analysis,we have worked out the scheme of WT-CFAR(Wavelet Transform Constant False Alarm) processor and scheme of WFTCS(Wavelet-coefficient Floating Threshold Clutter Suppressor).The WT-CFAR processor is used to reduce clutter of low frequency while the WFTCS can effectively eliminate clutter of high frequency.Based on the Nieman-Pilson Criterion,we have researched the selecting method of floating threshold, and given the CFAR probability coefficient -γof the wavelet coefficient of all scales, and we useγand the statistical mean-square-root value to construct the floating threshold of the wavelet coefficient of the above scales.Adopting the scheme,we can achieve the adaptive-control of wavelet threshold within certain probability of CFAR, and effectively restrain the composition of pinnacle of the clutter.No matter what kind of clutter,the wavelet coefficients always assume a normal distribution with mean zero,so this methed is adaptive for all kinds of radar clutter.Finally,the implentmenting model based on FPGA is discussed about the radar-clutter-processing algorithm.The results of the simulating experiments show the methods mentioned in this thesis are very effective and robust on the radar echo of sea,snow and land.