Direction of arrival estimation in uniform circular arrays composed of directional elements

We address the problem of estimating the azimuth and elevation angles of multiple far-field signals in uniform circular arrays (UCAs) composed of identical sensor elements. Unlike previous search-free approaches, which strongly rely on the omni-directional element assumption, the element pattern is modelled as direction dependent. We have developed the UCA-RARE algorithm, a new computationally efficient eigenstructure-based estimation method operating in beamspace. The algorithm allows estimation of the azimuth parameters of the source directions decoupled from the elevation parameters without employing exact knowledge of the element pattern in a search-free procedure. The corresponding elevation angles are estimated in a subsequent step using a closed-form algorithm.

1. Introduction

The uniform circular array (UCA) plays a significant role in many array processing applications due to its full 360º azimuth and 90º elevation angle coverage. The UCA structure is of particular importance since it requires a significantly smaller number of sensors to guarantee an unambiguous array manifold as compared to other highly structured planar arrays. Several direction-of-arrival (DOA) estimation methods have been proposed that can be applied to UCAs composed of omni-directional elements [1], [2], among which UCA-ESPRIT [1] attained increased attention due to its simple implementation and its ability to provide automatically paired azimuth and elevation angle esti-mates. However, this method strongly relies on the omnidirectional element assumption which is not only difficult to satisfy in real arrays but also limits the far-field phase modes that can be excited on an array at reasonable strength to certain frequencies of operation [3]. The use of directional element patterns, if carefully chosen, can overcome this limitations and open the way for wide-band operation of the UCA [4].