Research On Configuration Design And DOA Estimation Algorithm For Nested Array

Array signal processing has been widely applied in radar,communication systems and sonar.In the array signal processing,the sources angle estimation problem in targets localization,that is called the direction of arrival(DOA)estimation,is an important research topic.In recent years,nested arrays have been widely used in DOA estimation.Nested arrays enjoy more flexible layouts,reduced mutual coupling,larger array aperture,and enahnced degree of freedom(DOF)than traditional uniform arrays.Nested arrays enable to achieve targets localization with high performance and high resolution.The paper studies the arrays structure design and the corresponding DOA estimation algorithm in nested arrays.The selection of topic has theoretical significance and practical application value.The work of the thesis is summarized as follows:1)Enhanced nested linear array is designed and the corresponding one-dimensional local spectral search spatial smoothing multiple signal classification(SS-MUSIC)algorithm is proposed.Enhanced nested linear array enjoys more DOF and better DOA estimation performance compared to classical two-level nested linear array.The initial angle estimates of the virtual received signal of the enhanced nested linear array can be obtained by the Discrete Fourier Transform(DFT)method.The local spatial spectrum of SS-MUSIC algorithm can be constructed by utilizing the initial angle estimates,and hence,the proposed DOA estimation algorithm has low complexity.2)Widened nested linear array is designed and the corresponding sparse representation algorithm based on compressed over-complete dictionary is proposed.By decomposing and rearranging one of the sub-arrays of the generalized nested linear array,the widened nested linear array achieves the enhanced DOF and reduced mutual coupling.The virtual array of widened nested linear array is not complete continuous,and hence,DOA estimates can be obtained using sparse representation algorithm.However,the over-complete dictionary of the sparse representation algorithm contains the global spectrum,which leads to the computational burden.The DFT method is used to obtain the initial angle estimates,and then the over-complete dictionary is compressed by the initial estimates,which relieve the complexity.3)Enhanced nested linear array is extended to a two-dimensional(2D)structure.As a result,we design the enhanced nested planar array and propose the corresponding 2D local spectral search spatial smoothing MUSIC(SS-MUSIC)algorithm.Enhanced nested planar arrays enjoys more DOF and better DOA estimation performance compared to classical nested planar array.The initial angle estimates of the virtual received signal of the enhanced nested planar array can be obtained by the 2D DFT method.The local spatial spectrum of 2D SS-MUSIC algorithm can be constructed by utilizing the initial angle estimates,and hence,the proposed DOA estimation algorithm has low complexity.4)A new system structure which combines nested linear array and unmanned aerial vehicle(UAV)is proposed for emitters localization.A localization scheme is designed to obtain the two-dimensional DOA estimates of the emitters by rotating the UAV.The 2D DOA estimation problem is transformed into multiple 1D DOA estimations by the rotation characteristics.Specifically,the 1D DOA estimates is obtained by using a spatial smoothing based unitary estimation signal parameter via a rotational invariant technique(SS-UESPRIT)algorithm.SS-UESPRIT algorithm realizes the covariance matrix by means of a unitary matrix,so that the algorithm has low complexity.