Adaptive beamforming with near-instantaneous convergence for matched filter processing

We have investigated the implementation of relatively new broadband adaptive processing schemes in line array sonar systems with passive and active capabilities. We provide details of the investigation of adaptive beamforming schemes with and instantaneous convergence for matched filter processing. These schemes include a so-called memoryless generalised sidelobe canceller (GSC) beamformer and a coherent broadband adaptive algorithm that is based on a space-time statistic called the steered covariance matrix. Application results from synthetic and real experimental data show that the adaptive processing schemes provide continuous beam time series at the input of a matched filter. In addition, the adaptive beam time series have sufficient temporal coherence and correlate with the reference FM signals of active sonars, which indicates that the algorithms have achieved near-instantaneous convergence.