Fast Algorithm Of Joint Polarization And DOA Estimation For Irregular Array

Polarization-sensitive arrays can obtain polarization information and have higher system resolution than conventional arrays.Accordingly,we mainly study the algorithm of joint polarization and direction of arrival(DOA)estimation.However,due to the limitation for the antenna arrays layout in the actual environment,and the error existed in the process of fabrication,the geometric constitution of actual arrays is mostly irregular.Meanwhile,the four-dimensional search in the polarization domain and spatial domain is used by the joint polarization-DOA estimation algorithm,and requires a large amount of computation.Therefore,we mainly study the fast algorithm of joint polarization-DOA estimation for non-regular array,and the research conditions include incoherent sources and coherent sources.The main work of this paper is as follows:(1)Aiming at the large computational complexity for joint polarization-DOA estimation under uncoherent sources,the Quaternion Multiple Signal Classification(Q-MUSIC)algorithm and the rank loss principle dimension reduction MUSIC algorithm are studied.Both algorithms reduce the four-dimensional spectral peak search in polarization-DOA joint estimation to detached peak search of the spatial domain only and the polarization domain only,with reducing the amount of computation.Meanwhile,neither of the two algorithms is affected by array distribution.The theoretical analysis and simulation results show that the two kinds of dimensionality reduction methods have the advantage of reducing computation compared with the classical four-dimensional search method.(2)Aiming at the problem that the invalidation of subspace decomposition class algorithm caused by the rank loss in signal covariance matrix in coherent source environment.Firstly,the polarization smoothing is studied.By smoothing the subarrays of different polarization orientation patterns,the data covariance matrix is restored to full rank and the spatial parameters can be measured.Then,aiming at the disadvantage that the polarization smoothing algorithm cannot measure the polarization information,the generalized subspace fitting algorithm is studied to do search for signals in spatial-only and polarization-only domain.Finally,aiming at the shortcoming that the generalized subspace fitting algorithm is of large computation,the oblique projection algorithm is studied.The polarization vector of the signal can be directly projected and the polarization information is obtained,which decreases the computational complexity.(3)Aiming at the problem that polarization smoothing cannot measure the polarization information and the large amount of calculation in generalized subspace fitting,the method using virtual array is studied.Firstly,the polarized array element in the irregular array is virtualized as single-polarized uniform plane array,and the spatial parameters are searched by spatial smoothing decoherence.Then,virtualize the non-regular array as a uniform plane array composed of co-point dual-polarized elements,and search the polarization information after spatial smoothing decoherence.This method can virtualize any irregular array into a regular array,without formation influence.