Abstract
Direction finding techniques for uncorrelated or coherent signals have attracted thousands of scholars and researchers. It is proved here that uniform circular array (UCA) with an element on its center is able to conduct direction-of-arrival (DOA) estimation with eigenstructure-based methods in the case of coherence. This is achieved by dividing the whole array into two subarrays, each having four elements and both sharing the central sensor, and hence the subarray manifold is phase-delayed amplitude-weighted replica of another. By exploiting the property of subarray shift-invariance, it then becomes possible to create a smoothed array output covariance that is nonsingular, which can finally identify all direction of arrival despite their correlation. Furthermore, method proposed in our paper can be applied to all 6x2(q-1) UCA with an extra sensor on its center, while each subarray will always be consisted of four elements. In terms of uncorrelated signals, UCA with more degrees of freedom and higher array aperture than subarray can be well utilized to improve the performance of DOA estimation. Numerical results show that proposed method can resolve up to 6x2(q-1) uncorrelated signals and a pair of coherent signals along with relatively high accuracy.