Enhancing UCA-ESPRIT for non-circular sources

UCA-ESPRIT is a subspace-based algorithm that is applicable to uniform circular arrays (UCAs). It is a closed-form algorithm that provides automatically paired estimates of the azimuth and elevation angles of multiple farfield plane waves incident on the circular array. Work (M. Haardt and F. Romer, 2004) has shown how ESPRIT type algorithms can be enhanced when the signals incident on the array are non-circular signals such as BPSK or ASK. These enhancements were applicable only to arrays with centrosymmetric manifolds. This paper shows how UCAs can take advantage of the non-circular source enhancements (which include better estimator performance and the ability to resolve a higher number of sources) despite the fact that the UCA is not a centrosymmetric array.