Summary

International Symposium on Antennas and Propagation

2010

Session Number:3TA2

Session:

Number:3TA2-4

DOA Estimation Using Adaptive Beamspace EM Algorithm

Masahiro Sato,  Nobuyoshi Kikuma,  Hiroshi Hirayama,  Kunio Sakakibara,  

pp.-

Publication Date:2010/11/23

Online ISSN:2188-5079

DOI:10.34385/proc.52.3TA2-4

PDF download (147.7KB)

Summary:
To understand radio propagation structures and consider signal recovering techniques in mobile communications, it is most effective to estimate DOAs (directions of arrival) of individual incoming waves with array antennas. Also, in radar systems, it is required to discriminate the desired signal from interference [1]. Recently, DOA estimation using EM (expectation-maximization) algorithm [2] based on the maximum-likelihood (ML) approach receives much attention. It is because the EM algorithm remains stable in scenarios involving small numbers of snapshots, coherent signals and low SNR. However, the ML approach generally has the high computational complexity caused by optimization of the likelihood function. In addition, the convergence rate is slow, if the incoming waves are coherent and their DOAs are closely spaced. In this paper, we apply the beamspace processing using DCMP (directionally constrained minimization of power) criterion [3] to the EM algorithm for improving the convergence rate. Furthermore, we investigate the effect of the error of initial values in the beamspace EM algorithm on the DOA estimation. Through computer simulation, we show that the beamspace processing provides improved performance for DOA estimation.