Radar Complex Intermediate Frequency Signal Denosing Based on Convolutional Auto-encoder Network

In radar systems, target state features are commonly extracted from intermediate frequency signals. However, these signals often have a low signal-to-noise ratio due to noisy environments and limitations of the hardware. This can lead significant loss in performance during feature extraction. Therefore, improving is crucial for effective operation systems. To solve this problem, we developed deep learning-based method denoising paper. Our approach involves using an auto-encoder network remove unstructured noise recover original signal. During signal preprocessing stage, it important ensure that phase complex remains undistorted differences amplitudes do not negatively affect performance. achieve this, real imaginary parts separated subjected 0-1 normalization. The function then established based on correlation. numerical results demonstrate proposed outperforms other techniques terms mean square error