Low-Complexity DOA Estimation for Uniform Circular Arrays with Directional Sensors Using Reconfigurable Steering Vectors

Di Zhao,Zhongliang Deng,Weijie Tan
DOI: https://doi.org/10.1007/s00034-022-02188-z
2022-01-01
Abstract:In the direction-of-arrival (DOA) estimation methods for uniform circular arrays (UCAs) with directional sensors, the high computational complexity arises when achieving the estimation accuracy by leveraging all the sensors in the element space. We present a low-complexity sectorized DOA estimation scheme for directional UCAs. It models the radiation pattern within half-power bandwidth in a closed mathematical expression and thereby constructs reconfigurable sectorized steering vectors of an arbitrary angular sector by using the Hadamard cyclic vector with the Vandermonde structure. The DOA estimation over the azimuth plane throughout the element space is transformed to DOA estimates within multiple individual angular sectors. The new scheme comprises three stages, i.e., submatrix extraction, sectorized weighted beamforming and sector DOA estimation. The feasibility and effectiveness of the proposed scheme are validated by numerical simulations. The simulation results demonstrate that, compared to the previous sector-free DOA estimation methods based on the entire azimuth plane, our scheme has a low hardware and computational complexity under an acceptable estimation error.
What problem does this paper attempt to address?