Spatial Spectrum Estimation Algorithm based on Minimum Redundancy Linear Array: A Performance Analysis

Abstract

A performance analysis for minimum redundancy linear array (MRLA) in Moffet, A. T. (1968), one of optimal non- uniform linear array, is provided. After a brief back-ground introduction, we exam typical spatial spectrum estimation algorithms such as conventional multiple signal classification (MUSIC) in Schmit, R.O., (1986), maximum entropy MUSIC (MEM) in Burg, J.P. (1989), minimum norm MUSIC (MNM) in Ng, B. P. (1990), and minimum variance MUSIC (MVM) in Capon, J. (1969) and Hirakawa, M., et al, 2001, etc. with the direction of arrival (DOA) problem by comparing their statistical performances according to the theoretically perfect covariance, Cramer-Rao bound (CRB) in Stoica, P. et al, (1989). Based on the analysis, we further take the MUSIC, which is the most accurate candidate, as an example to do the comparison between MRLA and uniform linear array (ULA) on the statistical performance and resolution. By checking the results, we show that the MRLA beats the ULA on both the statistical accuracy and resolving power, moreover, we document the predominance for the MRLA on the sensor saving.