Reduced-Complexity Maximum Likelihood Direction-of-Arrival Estimation Based on Spatial Aliasing

Abstract

Spatial aliasing is an undesirable phenomenon that prevents the unique determination of the direction of arrival (DOA) of impinging signals in array signal processing. However, the characteristics of spatial aliasing that generates ambiguous DOAs can be also used to reduce the computational complexity in maximum-likelihood (ML) DOA estimation. This paper proposes a structural method to dramatically reduce the computational complexity of the ML DOA estimation using the spatial aliasing generated by a nested array structure with a doubly scaled aper- ture. An ML full grid search is computationally simplified by the highly compressed searching range and the small number of candidate values to be searched which are derived based on spatial aliasing. Performance analyses based on the theoretical bounds and computational complexity with computer simulations show that the proposed method requires an extremely reduced computational load compared to the conventional ML DOA estimation with a uniform linear array (ULA) while achieving the performance enhancement in the estimation accuracy due to the enlarged array aperture.