Ambiguities in Linear Direction-Finding Arrays

Abstract

Arrays used for direction finding (DF) are often plagued by ambiguities that can increase RMS error levels as well as generate occasional angle estimates that are totally wrong-these latter are often called “wild bearings.” A new analysis framework is introduced to understand how wild bearings are generated and where they are likely to occur. The analysis utilizes the squared correlation between array manifold vectors, leveraging prior work on the spatial distinguishability of cellphone users at an adaptive “smart antenna” basestation. It is shown that this metric, which was derived originally for adaptive beamforming, is also appropriate for DF algorithms such as Beamforming, Capon's method and MUSIC. Examples and simulation results are included that show the location and density of wild bearings for a simple linear array.