Convolution Neural Network Enhanced MIMO Array Radar Imaging Based on Back Projection Algorithm

The multiple-input-multiple-output (MIMO) array radar has plentiful application scenarios in the modern intelligent life. The back projection (BP) algorithm is widely used in MIMO array radar imaging for the superior imaging accuracy and generality of measurement configurations. However, the BP algorithm also bring the two-dimensional spatial distortion and background noise. To solve these problems, the U-Net network is introduced in MIMO array radar enhanced imaging in this paper. The experimental results of both random and fixed scatters show that the scatters are accurately reconstructed and positioned, the two-dimensional spatial distortion is corrected and the background noise is eliminated with the proposed method.