Extended Aperture Two Dimensional DOA Estimation

Realizing the two-dimensional angle estimation of the signal source to be measured is an important direction for array signal processing research.Aperture extension can improve the resolution of the array effectively,but it causes the cyclically ambiguous direction cosines.For the angle ambiguity problem,this paper presents two kinds of extended aperture direction-of-arrival estimation algorithms for non-uniform arrays: Non-uniform two-L-shaped array angle estimation algorithm based on propagator method and 2-D DOA estimation for non-uniform L-shaped array based on joint diagonalization.The main work can be summarized as follows:The first part proposes a high accuracy and low complexity DOA estimation algorithm for non-uniform two-L-shaped array with extended aperture.Firstly,the algorithm constructs an autocorrelation matrix based on the received signal vector.The linear operator based on propagation propagation method can obtain low-variance cyclically ambiguous DOA estimates and high-variance unambiguous DOA estimates.Secondly,using the high-variance unambiguous estimates as the reference,the low-variance ambiguous estimates with the smallest difference is selected as the corresponding low-variance unambiguous estimates.Finally,the simulation results show that the algorithm can get one-to-one angle estimation information,and reduce computational complexity effectively.In addition,the proposed algorithm can still perform accurate angle estimation,when elevation angles are within the actual mobile communication range.The partial array element spacing in the non-uniform array is larger than half wavelength to achieve the extended aperture,thus the algorithm has better estimation performance.In the second part,in order to solve the singularity problem of 2D DOA estimation with extended aperture,this part proposes a joint diagonalization 2D DOA estimation algorithm for non-uniform L-shaped array.Firstly,the algorithm obtains a large number of virtual array elements on the basis of Khatri-Rao product properties,thus obtaining a completely new array.The cross-correlation matrices with different delays are constructed according to the virtual array and the signal subspace is obtained.Secondly,the diagonal matrices containing direction cosine information are obtained by the operations applied on signal subspace.Then,the low-variance unambiguous DOA estimates are obtained through the joint diagonalization method and the ambiguity resolved method.Finally,simulation results verify that the algorithm can achieve accurate angle estimation in underdetermined case.The virtual array and the element spacing which is larger than half-wavelength extend aperture,thus the algorithm has better estimation performance.At the same time,the algorithm can also achieve accurate estimation when the DOA estimation matrix has the same diagonal elements.