Spatial-Smoothing-like Processing of array steering vectors for high resolution DOA estimation

This paper proposes spatial smoothing preprocessing (SSP)-like definition of the virtual array steering vector used in array mapping design process. Accuracy of direction-of-arrival (DOA) estimation is often deteriorated when incident signals are strongly cross-correlated, and SSP has been developed to reduce such signal correlation and then to accurately estimate DOAs of those signals. Application of SSP is limited to the array configuration where parallel shift of subarrays are considerable, like uniform linear array (ULA) or uniform rectangular array (URA). On the other hand, array interpolation processing is necessary to apply SSP to uniform circular array (UCA) which could provide DOA estimation in all azimuth angles with identical accuracy. This paper proposes a novel definition of the virtual steering vector as SSP-like form in array interpolation, and also the same concept is applied to element space (ES)-root-MUSIC.