FSF MUSIC for Joint DOA and Frequency Estimation and Its Performance Analysis

In this paper, we present a tree-structured frequency-space-frequency (FSF) multiple signal classification (MUSIC)-based algorithm for joint estimation of the directions of arrival (DOAs) and frequencies in wireless communication systems. The proposed approach is a novel twist of parameter estimation and filtering processes, in which two one-dimensional (1-D) frequency (F)- and one 1-D space (S)-MUSIC algorithms are employed-in a tree structure-to estimate the DOAs and frequencies, respectively. In between every other MUSIC algorithm, a temporal filtering process or a spatial beamforming process, implemented by a set of complementary projection matrices, is incorporated to partition the incoming rays to enhance the estimation accuracy, so that the incoming rays can be well resolved even with very close DOAs or frequencies, using the 1-D MUSIC algorithms. Also, with such a tree-structured estimation scheme, the estimated DOAs and frequencies are automatically paired without extra computational overhead. Furthermore, some statistical analyses of the undesired residue signals propagating between the 1-D MUSIC algorithms and the mean square errors of the parameter estimates are derived to provide further insights into the proposed approach. Simulations show that the new approach can provide comparable performance, but with reduced complexity compared with previous works, and that there is a close match between the derived analytic expressions and simulation results