Robust wideband beamforming algorithm based on frequency difference technique

This paper investigates a novel robust wideband beamforming algorithm based on the frequency-difference (FD) technique. While FD technique can maintain good frequency invariance (FI) of the beamformer, it introduces cross terms that can adversely affect performance. To reduce the cross terms, a projection preprocessing is implemented prior to FD operation exploiting the approximate orthogonality between steering vectors (SVs). Then, a robust beamformer is designed based on the estimated cross terms covariance matrix (CTCM) and SV of the signal of interest (SOI). The output of beamformer is an FD SOI. To restore the SOI, we design a filtering system and apply a simple diagonal loading beamforming to obtain the amplitude and phase of SOI, respectively. The simulation results have verified the effectiveness and superiority of the proposed algorithm in the presence of array errors.