On focusing preprocessor for broadband beamforming

The coherent signal-subspace approach to broadband adaptive beamforming involves a focusing preprocessor that aligns signal spaces at different frequencies to a common one by means of transformations and a narrowband beamformer following the preprocessor. The merits of the approach are decorrelation of multipath signals and partial adaptivity due to single-frequency weights. This paper deals with the optimality of the focused beamformer compared with that of conventional broadband beamformer. An analytical and empirical evaluation of different focusing transformations is presented