Fast DOA estimation algorithm using pseudo covariance matrix

Abstract

This paper presents a fast direction of arrival (DOA) estimation algorithm that can rapidly estimate the incidence angles of signals using a pseudo covariance matrix even under coherent signal conditions. The conventional multiple signal classification (MUSIC) algorithm requires hundreds of snapshots to produce a covariance matrix of signals, as such, it cannot perform a DOA estimation while acquiring this covariance matrix. In addition, the MUSIC algorithm cannot be used under rapidly changing or correlated signal conditions. In contrast, the proposed algorithm obtains the bearing response and directional spectrum after acquiring a pseudo covariance matrix for each snapshot. The incidence angles can then be exactly estimated by combining the bearing response and directional spectrum. Accordingly, unlike the MUSIC algorithm, since the proposed algorithm only uses a single snapshot to obtain the pseudo covariance matrix. It can rapidly estimate the DOAs of signals even when they are correlated.