A Method For Multidimensional Parameters Estimation Of Near-Field Sources In Impulsive Noise Environments

Direction of Arrival(DOA)estimation is a source localization technology and plays an important role in many fields.The tranditional DOA estimation algorithm assumes that the source is in the far field,the array element receives the plane wave.But when the source is located in the Fresnel region,the far field assumption will no longer be established at this time,so the traditional DOA estimation algorithm will no longer be useful.The current near source parameter estimation is mostly research under the background of Gaussian noise,but some of the noise in nature is often very impulsive and can not be described by Gaussian distribution.In the Non-Gaussian impulsive noise environment,the original mathmatical model based on Gaussian noise of near field parameter estimation algorithm will invalid.At present,the Fractional Lower Order Statistics has become a powerful tool for array signal processing algorithm in the impulsive noise environment.Based on the Propagator Method and Fractional Lower Order Statistics theory,is proposed based on the fractional lower order correlation and propagator method of MUSIC algorithm and based on the fractional lower order correlation and propagator method of ROOT MUSIC algorithm,respectively shorthand for PM-FLOM-MUSIC algorithm and PM-FLOM-ROOT-MUSIC algorithm.However,the algorithm based on fractional lower order statistics needs the prior knowledge of the pulse,and the prior knowledge is difficult to obtain in the practical application.So proposed based on Propagator Method and Nonlinear Compress Core Function Transform Correlation of ROOT MUSIC alogorithm,shorthand for PM-NCCFTC-ROOT-MUSIC algorithm.General near-field source parameter estimation is assumed the signal frequency is known,so the algorithm application will be limited.So,the thesis puts forward three dimensional parameters estimation algorithm which based on fractional lower order statistics and Nonlinear Compress Core Function Transform Correlation,the three dimensional parameters include near-field frequency,DOA and distance,shorthand for FLOM-ESPRIT algorithm and NCCFTC-ESPRIT algorithm.Due to the array parameter estimation algorithm based on sparse reconstruction theory than the traditional array parameter est-imation algorithm based on subspace with higher performance.Therefore,this thesis based on compression perception sparse reconstruction theory,combines fractional lower order statistics and Nonlinear Compress Core Function Transform Correlation,proposed based on fractional lower order statistics of the sparse vector quantization reconstruct near-field source parameter estimation algorithm,shorthand for FLOM-VEC,and proposed based on Nonlinear Compress Core Function Transform Correlation of the sparse vector quantization reconstruct near-field source parameter estimation algorithm,shorthand for NCCFTC-VEC.Finally,simulation results prove the effectiveness of the algorithm.