Algorithm Of Joint Polarization And DOA Estimation For Coherent Source

The field of array signal processing has developed rapidly in recent years.The super-resolution algorithm is an important field of array signal processing.There is currently an important research direction to improve the dimensionality of characteristic information to obtain better signal resolution.Polarization information is an important feature different from airspace inf ormation,so it is of great significance to estimate it.However,the coherent radiation source will greatly affect the performance of the algorithm,which will cause the rank deficiency of the covariance matrix of the data received by the array,and the t raditional MUSIC algorithms cannot make correct estimates.Therefore,this paper studies the polarization-DOA joint estimation algorithm of coherent radiation sources.The main work of this paper is as follows:(1)Aiming at the problem of rank deficiency of the data covariance matrix caused by coherent radiation sources,two smoothing-type decoherent estimation algorithms are studied.First,the spatial sm oothing algorithm is studied.This method constructs multiple sub-arrays with invariant spatial translation and rotation,and smooths the data covariance matrix of the sub-arrays to restore the rank of the matrix.Later,the polarization-spatial parameters can be adjusted by MUSIC and other algorithms.Estimation;For the spatial smoothing algorithm has st rict restrictions on the spatial structure of the array,the polarization smoothing algorithm is studied.This method smooths the data covariance matrix of the sub-arrays of different polarization component channels to achieve rank recovery.Find the spatial parameters.(2)Aiming at the requirements of the smoothing algorithm for the array space structure or the array element structure,the coherent source polarization-DOA estimation method of subspace fitting is studied.First,the generalized subspace fitting algorithm is studied.The method realizes the search of signal space parameters and polarization parameters through the identity of the array steering vector space and the signal space,and is suitable for arbitrary arrays.Then,in order to solve the problem of the large multi-dimensional search calculation of the generalized subspace fitting algorithm,a dimensionality reduction subspace fitting algorithm is proposed,which realizes the decoupling of the spatial information,namely the elevation angle and the azimuth angle,and further reduces the calculation amount.(3)Aiming at the problem that subspace fitting algorithms need to perform multi-dimensional search and the amount of calculation is large,the non-searching coherent source polarization-DOA estimation method is studied.First,the oblique projection algorithm is studied.This method construct s the projection factor through spatial information,and directly projects the polarization information of the signal,without the need for spectral peak search,and the calculation is low.Then,for the oblique projection algorithm to be used when the spa tial parameters are known,a two-dimensional PUMA algorithm is proposed.This method uses polynomial solving to estimate the spatial parameters of t he signal without performing spectral peak search,which constitutes a non-searching algorithm.Coherent source polarization-DOA estimation method.