Real-time channel data compression for improved software beamforming using microbeamforming with error compensation

GPU-based software beamforming has advantages over hardware-based beamforming because it is more programmable and complicated imaging algorithms can be efficiently implemented. However, the need for a high data transfer rate from the hardware front end to the software back end is a major technical challenge. To mitigate this issue, a micro-beamforming approach can be used to reduce the data size by pre-steering and summing the channel data over small sub-apertures. However, grating lobes and other artifacts may appear and degrade image quality. In this study, we propose a modified micro-beamforming approach to achieve real-time data compression while minimizing the artifacts. First, the RF data are converted to baseband and micro-beamforming is only applied to the amplitude data while the phase data is unchanged. Second, a compensation factor is used and applied to the subaperture amplitude beam sum at the software back end. Results show that the proposed approach can be implemented in real-time with minimal compression errors.