Some aspects of array signal processing

Abstract

The paper presents an overview of some important aspects of array signal processing in the context of narrowband applications. Inverse methods for solving the array signal processing problem are reviewed. Two case studies are presented, to illustrate the various issues discussed here. The first case study presents a comparison of various array signal processing algorithms; computer simulated data are used to conduct this study. The second study relates to the angle-of-arrival estimation problem in a low-angle tracking radar environment; the results presented for this case study are based on experimental data obtained by means of a sophisticated 32-element array system. Different kinds of nonuniformly spaced linear arrays are also reviewed, with particular attention given to minimum redundancy and nonredundant arrays. The tradeoffs; namely, array span and sampling gain that decide the choice of one non-uniform configuration over another, are discussed. To aid in this study, the well known Cramér-Rao lower bound (CRLB) and the newly developed Weiss-Weinstein lower bound (WWLB) are computed for the simple and yet informative case of a single source under different conditions. A novel method for the adaptive (online) control of the measurement space (and therefore the amount of information available for parameter estimation) is described. This method incorporates use of the CRLB as a component in online adjustment of the system. The paper also includes real-time processing considerations, and a review of calibration array techniques.