Adaptive Beamforming based on Convolutional U-net

In the current field of array signal processing, traditional beamforming algorithms face challenges such as limited signal enhancement capability and poor adaptability to high signal-to-noise ratio environments. To address these issues, in this study, we innovatively propose an adaptive beamforming strategy based on the U-net convolutional neural network. The aim of this method is to optimize the array receiver’s enhancement capability in the direction of the target signal while effectively suppressing interference signals and noise from other non-desired directions. By employing deep learning techniques, we accurately estimate the necessary parameters of the Minimum Variance Distortionless Response (MVDR) beamformer. Simulation experiments have validated that compared to traditional beamforming techniques, our proposed method not only demonstrates outstanding performance in both high and low signal-to-noise ratio scenarios but also significantly enhances the array signal processing performance while maintaining lower algorithm complexity. The research findings of this paper highlight the potential application of deep learning in the field of signal processing, particularly in signal enhancement and interference suppression, providing new perspectives and methodological foundations for future research endeavors.