Range-spread Target Detection In Compound Gaussian Clutter

With the wide application of high resolution radars,there are some new challenges in target detection under sea clutter background.Firstly,the amplitude statistical characteristics of high-resolution sea clutter appear serious tail compared with Rayleigh distribution model in the narrowband radar systems,and the Gaussian model is no longer suitable for describing heavy tailed sea clutter.Secondly,with the increase of radar resolution,the target is resolved into several scatterer centers on different range cells,which is called as the range-spread target or distributed target.The conventional point-like target detectors no longer meet the requirement.Thirdly,in the wideband radar systems,the traditional rank-1 target echo signal model is no longer suitable for describing the signals with Doppler spread caused by the translation and rotation of the target.Fourthly,besides the sea clutter in the traditional sense,for low-altitude flying targets,targets such as slow-moving ships on the sea surface are also regarded as clutter,and the suppression of such clutter is also one of the problems that need to be studied urgently in target detection under the background of sea clutter.From this point of view,the establishment of accurate models of sea clutter and target echo are the basis of target detection in sea clutter,and the optimal detector design under given sea clutter amplitude statistical characteristics is also crucial.Therefore,an adaptive subspace rangespread targets detector is proposed in compound Gaussian clutter with inverse Gaussian texture(CGIG)in this thesis firstly.Then,a new subspace signal detection scheme based on the rank-1 detector and multi-channel order statistics is also proposed.Finally,a point-like clutter map technique for eliminating slow-moving targets on the sea surface is proposed in combination with practical application problems.The main research results of this thesis can be summarized as follows:In Chapter 2,from the point of view of statistical modeling,the amplitude statistical characteristics of sea clutter are reviewed along with radar resolution varies from low to high.The amplitude distribution of sea clutter in compound Gaussian family with three specific textures is discussed in detail.The two step generalized likelihood ratio test(GLRT)detectors of range-spread target under different texture conditions are also reviewed.These detectors are the basis of the following subspace range-spread target detectors design.In Chapter 3,we investigate the adaptive subspace detection of range-spread target embedded in compound Gaussian clutter.The inverse Gaussian distribution is considered to describe the texture of the clutter in order to match the non-Gaussian characteristic of sea clutter.Moreover,the range-spread target's energy is assumed to be spread not only in the range dimension but also in the Doppler dimension,which is more suitable for the practical situation.Thus,the range-spread target is modeled as a subspace signal which describes the range-spread and Doppler-spread.Then,an adaptive detector for a subspace range-spread target is derived by using the two step GLRT.Finally,the detection performance of the proposed detector is verified by experiments based on simulated and measured data.The experimental results show that the detection performance of the proposed detector is better than that of the existing detectors.A new subspace signal detection scheme based on the rank-1 optimal detector and order statistics is studied in chapter 4,because the traditional subspace optimal detector proposed in chapter 3 is facing the difficulty of estimating the target steering matrix in a short coherent processing interval(CPI).To this end,a Doppler frequency components number dependent order statistics(DND-OS)detector is introduced,which takes the form of an average of maximum statistics under different normalized Doppler frequencies with a known number.If the number of Doppler component is unknown,the multi-channel order statistics(MCOS)detector is used.And the outputs from all detection channels are fused by an “or”fusion rule.Actually,the experimental results using simulated and measured data highlight that the proposed scheme can detect subspace signal efficiently.Chapter 5 draws lessons from the traditional clutter map processing idea,and records the interframe motion position and energy information of slow-moving target to establish the point-like clutter map through interframe iterating and updating.Based on the point-like clutter map,the slow-moving target can be eliminated.The simulated and measured data illustrate that the proposed method can eliminate the slow-moving target effectively.