Analysis and comparison of weighted subspace MUSIC and root-MUSIC using spatial smoothing techniques for smart antenna systems

This paper presents analysis and comparison of weighted subspace based MUSIC for the direction of arrival (DOA) using spatially smoothed forward (FS) only estimates, forward-backward smoothed (FBS) estimates of array correlation matrix. It is shown that for subspace technique like MUSIC, the FBS estimate techniques performs better as compared to FS estimates especially for correlated signals in multipath scenario. The ability of smoothing to obtain sharper peaks is applied under different estimates. The results are obtained for MUSIC and Root-MUSIC. It is observed that even for correlated signals MUSIC can be applied for DOA detection using above estimates. The resolution is also found improved.