Research On Modeling And The Detection Algorithm Of Non-IID Fluctuating Targets

The modern radar targets are significantly different from tranditional targets on geometry,material and the like.The targets' electromagnetic scattering are so complex and different that they will be non-independent and/or non-identically distributed(Non-IID)in the dimentions of space,time,frequency and polarization.However,the conventional radar cross section(RCS)models can not accurately describe the Non-IID character of the radar targets.Therefore,the performance analyses and designs for the Non-IID fluctuating targets will be lack of theoretical guides.As to the Non-IID fluctuating targets,the target RCS models and detections are studied in this thesis.The main contributions are as follows:1.One approach is proposed to model the RCS of the Non-IID fluctuating targets,which can contain conventional RCS models as special cases.The generalized RCS model can describe more accurately the non-identically distributed character and arbitrary correlation of the RCSs from different spatial directions,frequencies,polarizations and so on,solve the mismatch problem of conditional RCS models and provide a foundation for the performance analysis and system design.Meanwhile,the tri-parameter ?-? probability density function(PDF)is proposed to describe the target RCS,which can expand the kind of the RCS model.The new PDF breaks through the limit of target kinds described by the classical PDFs with one or two parameters.2.As to the sum of the independent and non-identically distributed(Non-ID)non-central Gamma random variables(RVs),the analytical PDF is derived.The approximately analytical PDFs of the sum of the independent and Non-ID Lognormal and ?-? RVs are given respectively.These PDFs provide the mathematic foundations for the computations and analyses of the detection performances.Then,the analytical performance expressions of the square law detection in the presence of independent and Non-ID non-central Gamma,Lognormal and ?-? fluctuating targets are derived respectively,and the truncated error upper-bounds of the performances are also obtained.The analytical performance expressions supplement the detection theory of the Non-IID fluctuating targets and provide theoretical guides for the performance analyses and system designs.3.One integral expression of the detection performance of the generalized constant false alarm ratio(CFAR)detector for fluctuating targets is provided,which can be used to solve the cumbersome computation problems of the CFAR performances for the Non-IID fluctuating targets.The theoretically analytical performance expressions of the cell averaging(CA)-CFAR and ordered statistics(OS)-CFAR in the presence of Non-IID Gamma fluctuating targets are derived respectively,and their truncated error upper-bounds of the theoretical performances are also obtained.The analytical performance expressions can solve the problems of the theoretical performance analyses of the CFAR detectors for the Non-IID fluctuating targets.4.Based on the mean powers' knowledge of the Non-IID fluctuating targets,one weithted detector with the searched weights is proposed,that can solve the problem of performance deterioration in the case of mismatched weights and improve the practicability of the detector.Based on the distribution parameters' knowledge of the Non-IID Gamma fluctuating targets,one bayesian detector with the searched parameters is proposed.This detector breaks through the constraint of the accurate requirement of the distribution parameters and solves the problem of performance deterioration of the bayesian detector under the condition of mismatched parameters.