A comparison of spatial interpolation and transformation matrices in the coherent signal subspace for bearing estimation of multiple broadband signals
Yuan-Hwang Chen, Feng-Peng Yu · The Journal of the Acoustical Society of America · 1991
The coherent signal-subspace (CSS) algorithm proposed by Wang and Kaveh [IEEE Trans. Acoust. Speech Signal Process. ASSP-33, 823–831 (1985)] for high-resolution directions-of-arrival (DOAs) of multiple coherent broadband signals focuses the signal subspaces at various frequencies into a common one through transformation matrices preprocessing. The preliminary angle estimates of DOAs using a spatial periodogram are often required to obtain the transformation matrices, and these angle estimates, if they are inappreciately selected especially in multigroup case, sometimes result in an estimation error in the transformation matrices and the performance of DOAs is thus degraded. Therefore, an interpolation preprocessing concept [Bienvenu et al., Proc. ICASSP, pp. 2799–2802 (1989)] was introduced to replace the transformation matrices without the need of initial angle estimates. The objective of this paper is to further compare the statistical characteristics and DOAs estimation performances of the two preprocessings coupled with CSS in the multigroup case. Computer simulations show that each of the two preprocessings has its advantages and disadvantages.