Research On Parameters Estimation For Near-field Source Localization

Source localization is an important research direction in array signal processing.It has a wide range of applications in sound source localization,wireless communication,wireless sensor networks,medicine,seismology and other fields.According to the distance between the source and the receiving array,the source location can be divided into the near-field source location and the far location location.In this paper,the narrow-band signal localization in near-field,near-field and far-Research,the main work is as follows:1)A near-field source localization method based on Rank Reduced(RARE)under uniform linear array is studied.The Rank Reduced MUltiple SIgnal Classification(RARE-MUSIC)algorithm,the Rare-Capon algorithm and the Rank Reduced Propagator Method(RARE-PM)algorithm are proposed using the idea of rank reduction.Based on the rank reduction method,the classical two-dimensional peak search class method can be simplified to several one-dimensional peak search,which avoids the huge computation amount caused by two-dimensional search and greatly reduces the computational complexity,and the method can Automatic pairing of the parameters is achieved,and the performance of the parameter estimation is very close to that of the two-dimensional search method,and the performance of the parameter estimation is high.2)A near-field source localization method based on Reduced-Dimension(RD)under uniform linear array is studied.The Reduced-Dimension MUltiple SIgnal Classification(RD-MUSIC)algorithm,the Reduced-Dimension Capon(RD-Capon)algorithm and the Reduced-Dimension Propagator Method(RD-PM)algorithm.Based on the dimensionality reduction method,the classical two-dimensional peak search class method can be simplified to one-dimensional local search,which avoids the huge calculation amount brought by the two-dimensional peak search,and can obtain the source parameter It is estimated that the parameters are automatically paired.Compared with the classical two-dimensional search method,the computational complexity is greatly reduced,and the computational complexity is further reduced compared with the reduced rank method.The performance of the parameter estimation is very close to that of the classical two-dimensional search method and the rank reduction Method,with higher parameter estimation performance.3)The method of signal source parameters in the far-field near-field mixing field under a uniform linear array is studied.Among them,based on the second-order statistics and the parallel factor algorithm,far-field and near-field sources can be identified in the mixed-field source with no more than one far-field signal source,and the parameter estimation of the source is obtained without The spectral peak search has the advantages of very low computational complexity and automatic pairing of parameter estimates.The reduced-rank MUSIC algorithm can identify far-field and near-field sources in any number of far-field near-field hybrid field sources,Field sources may be a variety of situations were analyzed,can get the source of the parameter estimates.Compared with the classical 2D-MUSIC algorithm,the computational complexity is greatly reduced,the parameters are automatically matched,and the performance of the parameter estimation is very close to that of the 2D-MUSIC algorithm,which has a higher accuracy of parameter estimation.