Study on the algorithm of real eigenvalue decompositions with a uniform circular array

The paper focuses on the improved 2-D MUSIC algorithm that operates in beamspace for DOA estimation with a uniform circular array (UCA). The algorithm requires only real eigenvalue decompositions to obtain noise subspace estimates, so that it has many advantages, including reduced amount of computation and enhanced precision, especially in a series of dual coherent sources scenarios. These excellent performances are verified by computer simulations.