Research On Underdetermined Direction-of-arrival And Polarization Estimation For Partially Polarized Signals

Traditional direction-of-arrival(DOA)estimation algorithms mostly require that the number of sources must be less than the number of physical sensors(overdetermined estimation),otherwise the algorithm will fail,which will inevitably lead to an increase in estimation costs.In addition to the DOA,polarization parameters are also important to describe the signal characteristics.However,most of the current literature can only achieve the estimation of the fully polarized signal,and it is generally aimed at the overdetermined situation,which has great limitations in practical applications.Therefore,it is necessary to propose an underdetermined DOA and polarization parameters estimation algorithm for partially polarized signals.In this paper,focusing on the underdetermined parameters estimation of partial polarized signals,the annihilation interpolation algorithm based on dual-polarized coprime array and the non-iterative subspace algorithm based on triple-polarized nested array are respectively presented.Simulation experiments prove the superior estimation performance of these two algorithms under low signal-to-noise ratio(SNR)and low snapshots.The main content is as follows.First,the structure and the mathematical model of the coprime array and nested array are analyzed,and the principle of improving the degree of freedom(DOF)of estimation is explained in detail.In addition,the specific motion characteristics of fully polarized signals and partially polarized signals are analyzed,and their joint representation in the polarization-spatial domain is given.Then,under the premise of receiving signals from the dual-polarized coprime array,several underdetermined partial polarized signal estimation methods are introduced,and the proposed annihilation interpolation algorithm is analyzed with emphasis.This algorithm reduces the dimensionality of the array output by adding sub-covariances,and interpolates the holes of the virtual uniform linear array with the help of an optimization problem constructed by the annihilation relationship,so that the DOF of estimation is further improved.Then combined with the obtained annihilation filter coefficient and noise variance through iterative solution to realize the underdetermined joint estimation of the DOA and polarization parameters of partially polarized signals.The simulation results show that the algorithm can more effectively improve the DOF of estimation and has a higher estimation performance.Finally,under the premise of receiving signals from the tri-polarization nested array,a method for deriving the Cramer-Row bound(CRB)is given,and a parameters estimation algorithm for underdetermined partial polarized signals based on non-iterative subspace is also introduced.The algorithm adopts sub-covariance addition to reduce the signal output dimension,and restores the rank deficiency problem caused by the second-order statistical signal by constructing a Toeplitz matrix that does not require multiple iterations,thus it has a smaller computational complexity.In addition,the algorithm also provides a nested array optimization method to realize the underdetermined estimation of the noise variance of the array.The simulation results show that the algorithm can effectively improve the DOF of signal estimation,and has smaller root mean square error compared with dual-polarized antennas,which further improves the estimation accuracy.